^^ My inputs may or may not reflect others’ situations. They can judge for themselves. The inputs I chose do result in net price spreads consistent with what the OP reported for MIT and Harvey Mudd. As I stated above, I’ve run these trials with many other inputs (and have reported the findings along with my assumptions on CC.)
I’m not really addressing the question of whether Barnard will necessarily offer more than Bates for any given candidate. I’m trying to address the topic question: Does the “meets full need” claim mean ANYTHING? If the 40-60 colleges that make this claim generally offer better n-b aid (and lower net prices) than similar colleges that don’t make the claim, then it isn’t a meaningless claim (notwithstanding the fact that a few colleges may be playing fast and loose with the claim). It is a more or less testable claim (unlike, say, “more prestigious” or “better for pre-med”) but I don’t see how one can test it without running numbers.
Again, absolutely, you do need to run your own NPC estimates. However, it isn’t practical to run the NPCs on hundreds of colleges. The problem cries out for “satisficing” (https://en.wikipedia.org/wiki/Satisficing). So, which heuristics are most useful for constraining the initial large search space? IMO it is useful for the CC community to assess the “meets full need” claim (including the boundaries where it tends to break down), along with other more or less testable claims, so that families don’t focus their searches only on famous/familiar/nearby colleges that may not offer the best combination of quality and value.