This is very accurate. As I mentioned before, the marginal delay impact of MDBF differences between fleets is very small, and only becomes worth considering when the relative difference between the MDBF of two fleets is over ~150K.
Besides the ability to handle crowds, there are some other factors, like electronic vs. non-electronic signage and shop familiarity with car classes that go into the decision-making process. However, the main goal is usually to maximize throughput, with all other factors being secondary, subject to cost considerations. A good example was the decision to place the R62As on the 6, rather than the 4, in order to not "contaminate" the entire Lex express service with the empirically-demonstrated increased dwell times of the R62A fleet in areas of high station crowding.
I'm sure that this has been repeated before, but the term "overcrowding/insufficient capacity", while used as a delay "cause", is more accurately a definition of a symptom of a delay. Many operational factors, including the relative impacts of car assignments, can have an impact on this "bucket" of delays. I'm sure you (RR503) know that, but I think it deserves explicit clarification for others. To put it simply, overcrowding is often a symptom of other operational issues, rather than a discrete cause of delay.