That's a broad discussion. Using your example, however, I'm inclined to disagree. Smaller units are generally preferable to fewer, larger units from a militaristic view. Smaller units are more affordable to produce from the outset, and since an investment is spread across several units (instead of a single) expensive unit, relatively speaking, any losses incurred will also be spread, and perhaps ineffective, since the attacks itself will be spread. It will also depend on the role of the spacecraft in question. I'll agree on the 'manned' part. To the overall point of effectiveness, however, I don't think that can be established. Additionally, anything set in the far future will be difficult to analyse. I can say with certainty that mechs will always be garbage. Giant mechs, at least. Man-sized, remote controlled or AI controlled mechs could be useful in a few instances. Say, for example, in clearing mine fields or in attempting to disarm explosives.