I'm not totally sure where things are now, but based on what I know, the current system is bad because 1) you can get elite level (if high risk) talent for a small cost (compare $5m to when JD paid like $75m for Bug) and 2) the tie-breaker for prospects signing was whichever team is "best", which is an obvious advantage to better teams
The old system was better, but it had issues too. It got to the point where you basically had half the league competing for IFAs on alternate years - of course you had people like Wooly taking it too far with fuckshit like signing major leaguers to contracts with alternating heavy years, like $3m, $25m, $3m, $25m, etc. This was better than current because 1) to get the high level talent, you had to pay a huge premium and 2) if you wanted a prospect bad enough, you could get him.
The biggest problem with abolishing it altogether is balancing talent levels. I remember I had to do A TON of different sims with different feeder league levels to find the right amount coming through, and I don't see how you can do it without spending a lot of time doing that testing. There's some incentive to tank more, but I don't really think any more than with the old system - if anything, you could basically plan to spend every cent you had for one given year on prospects, fill your major leagues with min contract guys, and get 1-1 and the top prospect in IFA.
Abolishing it also creates one clean path of talent generation (possibly with discoveries) which means version control of OOTP is less important and talent levels overall remain more consistent throughout seasons (but again this is only accomplished through significant testing). Drafts become more interesting since the talent level is higher - this is a while ago but I believe you tend to see 0-2 GOATs, then a batch of All-Star level talent, then a huge group of starting level talent that pushes multiple rounds, then you get ur lottery picks in slightly later rounds followed by the junk.
Obviously I'm not involved anymore so my opinion should be treated as meaningless, but I did have a strong opinion here and figured it was worth sharing. No IFA can be a hassle, but if monitored and tested correctly is the smoothest long-term solution to variable talent levels. Some people really love IFA and building through multiple avenues, so the old system accomplished that better than the current.