I would hope that management takes a lump of the responsibility for the decision making process and what they set-out to make the development team go through and end-up with.
Managing a studio in the Videogame industry is a hard thing. Typically when a studio gets funding cuts, it's up to the powers at be decide on how to reshuffle funds, I've never been in those meetings, but the pattern seems to be leaving a core group that can keep the studio running as best it could. Layoffs would go through all departments, including management, if that's what you mean by "taking lumps".
If you're trying to blame management for not being able to schedule something as complicated as a videogame with a limited schedule, I'd say, it's not easy. Videogames are both art and technical, mixed with the fact that it's very easy for scope creep, or "risk" that often comes to fuitition in terms of bugs or things requiring rework, makes scheduling very difficult.
In my tenure in videogames, either you have a formulaic project, like a sports game for example, the features and risks are fairly known, often swappable from the last years game (AI/Rendering/etc), the only big question mark is on generation changes, as you are reliant on external factors (hardware/API's/Tools/Documentation/etc...)... and even so you have massive failures. For example, there were years at EA where NHL was terrible due to features not working out, or NBA trying to implement new stuff that just sucked.
The other side is a studio trying to either do something new, or trying to prove their studio is awesome (which garners work and more freedom with publishers). This situation sucks, becuase the powers at be in your studio is trying their best to convince the people with the money that the game they are pitching, or in development is going to be awesome, so keep sending money (or let us work on it a bit longer than we origional estimated). This is where vertical slices come in, throw everything cool into a short demo, estimating gameplay, and also the graphical fidelity that we expect to hit. Eventually you have to make an actual game, but 1 virtual slice + 6 more do not make a complete game, and the coming together of everything is incredibly difficult to do, ontime, and on budget. So much unknown, risk, and estimating work, QA, regressions, polish, more bugs, etc... is hard.
A bit more on topic with regards to projects I worked on.
The first [Prototype], the studio was pushing Activition hard for more budget, we knew we had something that could be incredibly fun. Early on, the uppers indicated that activation was going to give us X million in funding, we were all like YAY! Throughout the process, activition sent our game code to some reviewer company, and initial indications were that we would hit high 80's I believe, activition was pretty ok with that, now they started commiting on marketing, and we had sales predictions. Marketing started doing their thing, and it was incredibly exciting to see fan feedback, and excitement to trailers and such. But in all honesty myself, and likely others knew that high 80's was probably too high, as we were completing the game, I knew there were some aspects that were awesome, but other aspects that sucked (like two boss battles in particular, also the technology wasn't quite up to some other games coming to market, even in our genre, remember the prototype vs Infamous?). This is around the time where Activition really started pushing for "few games" that "sell a LOT", and metacritic seemed to be their indicator (getting a new IP green lit from a publisher is incredibly hard to begin with). I recall finalling that game to be quite stressful, because Activition started pulling back, as they started seeing the more final code. When it finally came out, intial reviews were fantastic (not sure if something shady happened there,) but eventually it leveled off lower than original estimates.
When I worked on Mod Nation Racers, we knew we had something revolutionary going on, I mean kart racing on the playstation where you could build your own tracks?! Sony is an incredible publisher, very patient, and let us do our thing, we didn't quite meet all our deadlines, but we were still incredibly excited. However, some things we knew were going to hinder the reviews, things like load times, framerate, weapon balance mostly. It was pretty dishartening to see the sales be so poor. For me in particular I was disappointed that some parts that I really enjoyed and helped polish didn't get as much love (seemingly) from the community. I really enjoyed time trail mode, certain tracks really are awesome for shaving off miliseconds, and with competitive ghosts, it was a lot of fun getting the drifts just right to just edge out your friends ghost) Anyways, reviews were ok, sales were bad, little big planet on the other hand was very well received... guess what the studio was kind of forced to make next? In the end, it was disheartening seeing the community not give it a chance due to poor reviews that focused on some of the annoyances. Yes load times sucked, but online was actually awesome, the track editor was awesome, and the racing/drifting/physics model was quite good once you got used to it.
Sleeping Dogs, or True Crime Hong Kong :) We knew we didn't have the budget as GTA, so we focused on what we felt we could do better. Again, Activision, the demos were good, tug development was starting to go a bit long as we were trying to finish off some bits and pieces (like the story missions lol, though we were close!). Eventually the whole project got canned, to later be scouped up by Square Enix. I think we all expected to be in the 80's. but for that game, since it was back from the dead, I think the most positive thing was the positive community reaction. :)