Self-selection, in the past, when you had an extra tenth of a second to react, you didn't have to pick up the spin as quickly and you could have a longer swing. So guys with better eyesight and faster bat speed didn't have as large of a competitive advantage, which mean players with other attributes, power, speed, etc. could win jobs.
As FB velocity increases, some batters respond by developing more compact swings (faster bat speed, waiting longer on pitches) which was helped by video analysis of swing mechanics (easier to note flaws). As players trained better and got stronger, they generated more power out of those quicker strokes, which made more hitters willing to make that adjustment (and higher velocity FBs means you can let the pitch create exit velocity).
But it also provided a competitive advantage to hitters who could "see" better, which is both eyesight but also brain processing speed - not sure how much of that can be taught or trained.
You can see this in games now, it used to be that a 97-98 MPH FB was unhittable, now if you get one over the plate, a lot of hitters can take you downtown (see Chapman in the WS). So while minimum acceptable velocity has increased slightly, the marginal value of additional velocity has declined relative to command, movement and the ability to change speeds and keep hitters off-balance (which slows their reaction time).
To me, the great equalizer is the changeup (and the splitter, which has similar attributes) because you can't pick up the spin out of the pitcher's hand unlike the FB, slider and curve or even a cutter. A good changeup makes a marginal FB look much faster (see Hellickson when he had it working) because the hitter hesitates to see if it's a change or FB. If I was running an organization, I'd make teaching the change my #1 priority, and I'd scout for it in college pitchers (I doubt many HS pitchers have plus changeups).
