Hockey and I think about this in similar ways.
I haven't found a ton of value in spending $20M in both college and HS because while you're increasing your pool options, you're still stuck with a static amount of picks. So you're investing $40M in taking, at best, probably 4 or 5 guys who might make your ML roster one day if you have a couple of comp picks? Pass.
The value in paying the full $20M is the accuracy of the ratings. This obviously is helpful in the draft, but also if you're keeping a spreadsheet of the players from the draft pool each year, it's useful data to go back to when making trade proposals, particularly if you believe in $0M for Advanced Scouting, like many here do.
Be aware, 1 or 2 guys from whichever college or HS you have $0M scouting against still could show up in your Top 50. Typically I'll notice splits looking too good to be true and realize the guy is 18 years old (I'm a college scouter) and move him down my list. I find that those players who make it through are usually pretty good but you're taking a wild/blind chance given the inaccuracy of those ratings.