Steam is over-earning (as we call it in the investment community).
First, the only reason Steam is charging less than 30% (on some games), is that Epic came around. They would continue with 30% for everything if Epic wasn't there, without a doubt. Epic coming around is already a massive win for the gaming industry, due to more options and increased competition in the marketplace. Before, if a publisher/developer wanted to distribute a game, they could either do it directly (which carries a lot of frontloaded investments in the platform) or go to Steam and pay the egregious take rate due to their monopolistic position.
Steam is over-earning in the sense that they capture more than their fair share of the industry profit pool, given the value they provide. Generally speaking, every industry can be simplified as: suppliers -> distributors -> customers, this is called the value chain and can often be extended. Steam falls in the distribution category and is an aggregator of supply (games) with a direct relationship with the customer and sells games at close to 0 marginal cost (digital distribution cost is minimal, only small % of payments).
Steam is probably ~60-70% operating margins, because there is minimal marginal costs of distributing a game. Most of the costs they have are fixed (servers/infrastructure/some R&D). The only reason why they wouldn't have even more than 60-70% margins is that they are using some of the margin to re-invest in sales and marketing (gift cards etc.), to increase sales and profits even further, but they don't have to. Stop thinking that 60-70% margins are necessary for Steam to provide the "features" that Steam-players love so much. Those are mostly investments made at one point in time, not ongoing expenses. There is a "one-time" cost for developing the feature (R&D/personnel costs), with very limited ongoing costs. As a result, there is no need for 60-70% margins to maintain/or develop more features over time.
60-70% operating margins are relatively common for marketplaces when they have gained a monopolistic position and scale. What do they provide to the value chain to "deserve" that proportion of industry profits? For developers/publishers, they primarily provide two things: (i) access to their customer base and (ii) infrastructure (features/servers/payments ecosystem). When the PC market was small (due to disregard from publishers due to high risk of piracy) and infrastructure costs were high, for most publishers it didn't make sense to invest in own infrastructure. The return on capital on that would often be lower than just distributing the games on Steam. However, that has changed. Infrastructure costs have come down significantly (due to large scale build-out of datacenters, which has made it more accessible and relatively cheap) and costs to implement features have also come down (a lot of these are commodities). As a result, it is a more viable option to distribute direct-to-consumer (the holy grail for publishers, and suppliers in any industry). However, not all publishers can do this, as there is still (i) infrastructure investments to be made and (ii) relationship with customers (players) need to be strong enough for customers to want to come directly. There are a few players where that is the case (i) Battle.net (Blizzard), (ii) UPlay (Ubisoft), (iii) Origins (EA) and (iv) Rockstar Launcher (Take-Two). For them, the 30% is way too high. Why should they give away an enormous part of the profit pool (while carrying virtually all the cost), just for distribution that they can do themselves? There is definitely a "take rate" that Steam would charge where these would come back and lean into the platform, but I suspect this is well below 10%. If Blizzard, Ubisoft and EA don't use EPIC at 12%, you can be sure that it will be lower than that. For the relationship between Steam and large publishers, Steam should charge ~7-8% (just an example) and as a result, they are over-earning by an incredible amount (30% vs. 7-8%) relative to publishers. At 7-8% take-rates, Steam would still make a good profit (they would probably need to cut down on sales & marketing). There are many market-places that charge these types of take rates that still make a good living.
The crux is, while they should earn 7-8% for the large publishers, there are also sub-scale publishers (where necessary direct-to-consumer investments are too high to justify), where distribution is worth a lot more. For them, there has (until now) been no other option than Steam, and they have had to pay the 30% take rate for a long time. What should they do? Make console games instead? No, there they pay the 30% too, but there the distribution is worth significantly more "value capture" within the value chain.
Console platform providers can much more easily justify a 30% take-rate platform fee: (i) upfront investments (multiple years of R&D and development of the console itself), (ii) low gross profits of selling the consoles (even though gross profit margin can now be positive, doesn't mean it is good) and (iii) no other option if a publisher wants to reach the "captive audience".
It doesn't matter that gross profit is now "positive" it is still incredibly low compared to other Microsoft products (typically 90%+ in software), there is no way in hell that they would produce consoles and sell, if it wasn't for the platform fees they then can generate over the life of the console. Their alternative use of capital is worth much more than that.
Now, from the publisher/developer point of view in terms of market-place distribution, there has been three viable options (i) PlayStation (Sony), (ii) Xbox (Microsoft) and (iii) Steam (Valve). All of them has charged the same 30% even though Steam has provided much less "value" than the other two. Because Steam has had 30% margins, publishers have typically been indifferent between Playstation/Xbox/Steam when developing games. But if the PC take rates come down from the 30% (which it already is doing), you can be sure that publishers/developers will prioritize PC as a platform, relative to Playstation/Xbox. Both Sony/Microsoft will be scared of this (and should be very scared). They should also be scared of cloud gaming/streaming platforms, which provides more competition on distribution for console equivalent gaming. Stadia for example, will definitely have lower take rates than Playstation/Xbox (I would guess somewhere in the 15%-20% range) and then there is "rumors" that Amazon will participate at some point, Tencent will do something at some point - and suddenly, you have a competitive marketplace (instead of a cozy duopoly). Microsoft/Sony's take rates will already be under pressure with their 30% when PC take rates go down, but will be under even more pressure down the line. Wouldn't be surprised if we in 10 years time see take rates in the 20%-ish range.
So for publishers, Epic matters a lot, they have been the key catalyst in providing an alternative to Steam, who subsequently cut rates. They have already won, to some extent. The "dream" scenario is Epic new player with low take rates -> force Steam to cut rates -> force Xbox/Playstation to cut rates -> race to the bottom where it makes economic sense for all. It is therefore critical for publishers to keep Epic alive (and make them successful), because it will benefit publishers/developers for the entire future of the industry. Once take rates are going down across the board, the publisher/developer captures more profits and they can in turn re-invest those profits in either (lower prices of games) or keep the profits (probably unsustainable). In 10 years time, I not only predict that F2P will continue to grow and dominate, but that the prices that has basically been the same for 20 years, will come down as the industry becomes 100% digital and take rates come down. The only reason it is not coming down right now is (i) physical/retail sales requires same prices as digital (otherwise it would die immediately) and (ii) below $60 for physical/retail doesn't make sense economically. Once those two barriers are removed (and take rates come down), the industry would benefit from lower pricing, to try to make gaming even more mainstream than now.
Steam could capitulate and go towards take-rates where they earn fair value capture - but why should they? They would forego a lot of profits for a long time. The risk for them is that they will become obsolete over time as other alternatives emerge and allow them to grow/become better. Steam could have killed Epic before they even started with matching the take rates of Epic (would be very costly in terms of profits), but there is no way that Epic would even have existed then. It will likely be a slow burn, because it is too painful for Steam to cut take rates right away, over-time that will allow Epic to grow bigger and improve their product (at some point, in a few years, the product offering with likely be equivalent), but they definitely could do it. I suspect that the only reason they are keeping their take rates for now (20-30%), is that they hope that Epic will run out of money and eventually lose, so that Steam can keep their take rates (it will be very difficult to raise them again if they do, that's why they are doing the "middle-ground" at the moment).
In my view, what Steam "should" do, I if I was a board member (it is very difficult) is to (i) lower take rates (accept that you are over-earning, had a good run, but in order to maintain dominance, need to cut it), (ii) try to price discriminate even further (what they do currently with offering certain take rates (i.e. the 20% above $50m)) -> because the value Steam provides is hugely different to different publishers/developers (worth 7-8% for big publishers but 30% for small ones)) and (iii) develop more games to make their supply unique and differentiated again (that's how they started). Then they could become a formidable (unpenetrable) platform again.
For Epic, what they are doing makes total sense. The Steam userbase has a high lock-in and provides better services. It is too costly (too much of an investment) to replicate all features immediately, because even then, the switching costs from Steam is high (library, friends, ecosystem). They need something else to win and get an audience, which is unique supply (exclusives), that's the only way they can realistically compete to earn enough money, that they can then reinvest in making the platform better. Epic has very smart investors, such as KKR (with large amount of capital) and the opportunity is to great to not go for it.
In the end, content/games matter the most, not the storefront, the publishers/developers are playing the long-game here and want to earn their "fair share" of profits over the long run.