Stop beating your own straw man.
When have I ever said velocity is unimportant, that's like saying speed or power for a position player is unimportant.
Velocity is merely part of a pitcher's arsenal.
Velocity itself doesn't give a huge advantage, there are numerous pitchers who throw in the mid-90s who never get out of AA ball.
Velocity, like power, gives you a larger margin of error, a power hitter can strike out a lot, a high velocity pitcher doesn't need great command.
Where velocity matters is at the extremes, under 90 MPH there are few effective ML pitchers, over 95 for a starter or near 100 for a RP, you can get away with more "sins." But very few pitchers have high velocity, most SPs average between 90-94, most RPs 92-97. And high velocity pitchers tend to break down since they throw mostly fastballs, which is correlated with arm injuries.
So most prospects will fall, or be projected (as 16 or 18 year olds) to fall within the "normal" range.
"Natural selection" will mean those at the bottom of the velocity range will have better command and secondary pitches (or they'll fall out of the sample), those at the top of the range will develop other pitching skills as they age and lose velocity (or they'll fall out of the sample).
All other things held equal (that is, velocity isn't due to overpitching or other factors that limit pitchability), prospects with higher velocity will be valued more than those with lower velocity, but even among prospects, those with lower velocity will tend to be more polished or they won't get drafted in the first place (if Eshelman had Velasquez's skill profile other than velocity, would he even be considered a prospect?). So when we look at the sample of legitimate prospects, those without the right combination of skills have already been weeded out.
Because velocity is easier to measure, it will tend to be overpriced in the market, other attributes are harder to scout and thus provide bargains if you have an above average FO - and if you don't, you're screwed in any case.
Or to put it simply, higher velocity pitchers have higher ceilings, but not necessarily the same probabilities of reaching that ceiling. So a lower velocity pitcher may well have the same "expected value." Now you may want to balance your expected means and variance, in that you don't care about prospects with ceilings below a certain level, no matter how high the probability of hitting that level, and want some prospects with high ceilings and low probabilities of hitting those ceilings since there's a ML roster constraint (5 starters, 7 RP, what your 13th pitcher does has little value other than organizational depth). But that's an optimal portfolio problem.
