It's a matter of organizational philosophy, draft is more flexible than IFA, and ADV is still mostly useless. Since the most recent wave of updates, I personally observe that budgets relate exclusively to volume moreso than accuracy. Budgeting 6 or 16 is equally inaccurate; 16 simply gives you a larger pool.
Draft is flexible based on team context, specifically # of type-A and type-B free agents you are schedule to lose or to bring in, etc.. In seasons where you are losing free agent players you will have more picks in the top 75 (the most I've ever picked in the top 100 is 12), therefore you can re-distribute player payroll budget units into scouting for those specific seasons. If you're drafting 12 times in the top 100, you're going to have booms and busts alike. Because every budget includes wildly inaccurate scouting, every pick is a guess. I give myself the most guesses with the largest pool. When I don't have a lot of guesses, I shorten my pool. My teams draft better players because of the effects of trickle-down rather than accuracy
For these reasons, in my opinion it is smarter to balance your HS and COL as 10/10 rather than 20/0 or 0/20. In seasons which you draft more often you can raise to 14/14 (+8 units total), as opposed to 20/4 or 4/20 (+4 units total), and in seasons where you want to attack free agency you can go down to 6/6. Having a wider range (12 units thru 28 units) is more flexible. In similar fashion, draft/IFA scouting is +/- 4 but with prospect you can go from 6 to 20, and if you aren't doing IFA there is no reason to have more than 8 or 10 in prospect, which gives you more money for big league free agents and/or better draft scouting.
IFA is a heavier commitment and much riskier. You basically must commit 15+ to IFA scouting and commit 20 into prospect, plus transfer penalties. Even with that you might sign a bust, or you might not sign anybody at all. Draft >>> IFA