The average team is far too tentative about the draft. The way the system is setup, the best talent should get distributed to the worst teams, and eventually some form of competitive balance is created. Sadly, things don't always work out the way. A solid portion of the players in the draft get taken far later than they should thanks to signability issues. Teams are unwilling to grab the best talent because they aren't willing to give the best talent the money that they demand. By taking advantage of the strong players that fall to late rounds, a team can fill their farm system with talent. All they have to do is divert a few million dollars from the big league team towards the draft.
Look at the Dodgers fifth round pick this year, Kyle Blair, as an example. Blair was named the 29th best player in the draft by Kevin Goldstein of Baseball Prospectus. If teams simply drafted players based on talent, Blair would have gone either late in the first round, or early in the supplemental round. However, Blair wanted, gasp, a million dollars so he slipped all the way to the fifth round, where thankfully the Dodgers snapped him up.
What's strange to me is that the teams that let Blair pass them by wouldn't blink at throwing that same money to a backup catcher or a bad middle reliever. Over the course of a year how much value would these players provide? Half a win? Meanwhile, if a guy you draft becomes even a mediocre player he can provide fan more value over the span of six years.
Paying these talented players the money they demand is almost guaranteed to payoff in the long run. In today's market, picking up a mediocre starting pitcher for six years could cost 60 to 70 million dollars. If you get the same pitcher through the draft, he'll only cost you about 20 million for the six years he is under the team control. Not only that, you'll be getting the player's prime years instead of the downside of his career. Every starting pitcher you hand those million dollars to has the opportunity to save you 40 million. In other words, you can watch 34 picks that you give one million dollars not pan out, and if that 35th guy becomes a decent starter, you've still saved five million dollars. Since I have to believe that a player with first round talent has more than a two or three percent chance to pan out, this can't hardly even be classified as a gamble.
Now I realize that this isn't a totally realistic plan. If a G.M. flushes 30 million dollars of his owners money down the toilet, he's not going to have a job any more. Because a G.M. has almost zero job security, he'd much rather spend that million dollars on something that will provide some sort of visible benefit rather than something than a long term plan that is almost guaranteed to pay off but can look very bad in the short term. Even so, the simple plan of just drafting the best player available is sees surprisingly little use, and any team that takes advantage of it can quickly amass an amazing set of talent. All they have to do is live with having a two or three fewer professional hitters on the bench.