I'm not saying to ignore harsh realities, but that having fun is also an important aspect. As you said, you don't want a purely financial game, you also want player development even though it might not yield the best ROI.
Speaking of ROI, I would advise that every time you have some cash to put into your arena, you calculate the current ROI of each of the different seats and expand the one(s) that currently have the highest ROI (instead of just starting with the luxury boxes and courtside seats). You should keep in mind the final layout of the arena, which is indeed something like Wolph suggested (50/500/5000/15000) The boxes and courtsides are fixed at 50/500 maximum, and depending on your view you can build a couple thousand less or more lower tier seats (I went for 7000) and fill up with bleachers up to 20.000 total seats. You won't reach those numbers any time soon (takes multiple seasons), so you can just build the most profitable seats ROI-wise for the time being.
About transfer prices, there's indeed a large number of overpiced players on the market, and managers do bid irrationally because they're unknowing or impatient (not everyone has time to hunt for the optimal buy, and just want a player right now, the instant gratification generation so to say). So patience pays off on the market. The transfer estimates generated by the game are not very accurate (though it does give you a slight idea if you don't know any better way to determine value), because it does not account for the value of individual skills (it just looks at age, potential, salary and TSP in recent period).
Speaking of value of skills, OD is indeed generally accepted as most valuable, followed by IS. Besides that there's difference in opinion as different player and team builds evolve, but something to take into account is the training time required for skills. OD trains slow, DR/HA/JS/RB train fast so are generally less valuable. This is especially the case for out of position skills (OD/PA also for bigs, IS/ID for guards). Also the alternatives available for a skill weighs in. A guard with low OD is bad at defending, but if he has low JS but high IS he can still score (you're just limited in offensive schemes).