That sounds like an argument to hide the potential rating, not delete it.
If people want ZERO potential rating in the game (i.e. not hidden, but it doesn't exist at all) how would you cap player ability to make sure every player didn't reach 99 and do it realistically?
In real life, every player has a limited amount of potential. Not all WR are Jerry Rice or Drew Pearson. Not all QB are Dan Marino. Not all MLBs are Mike Singletary.
So something needs to limit players and in a computer game, that requires a number since that's what computer games work from.
As far as the idea I've seen saying stats should = progression, I don't find that realistic. A WR doesn't get better at catching balls by catching more balls in a game. A WR gets better at catching balls which then means he's better at catching them in the game. Stats driving progression is doing it backwards imo.
Plus, just because a player is struggling on the field doesn't mean he's not learning off it. It just means he's not starting caliber yet. Yet the stats = progression idea would make this guy suck and lower is ratings, making him suck worse, which lowers ratings worse, etc.
Madden needs a more in-depth system for player development across the board.
Some ideas on what this could be:
-Yes, I do think there should be potential ratings, one for each and every skill. Make them hidden, however, and available - as a GUESS - based on scouting you hire and time the player has been in the league. Rookies should almost always have "fog" around them - vets (i.e. players who have played a lot, use downs played stat) should almost always be easier to scout.
-Dynamic development. Injuries should have a chance to damage potential. A guy with bad knees might never be what he could have been if he never got hurt or before the injury. It should go the other way too, guys should "get it" sometimes and boost their upside in skills. Make it vary based on age (mental should "click" after some years most often while raw physical skills should "click" on younger players more often)
Likewise, aging should impact this. Older players should lose potential faster in physical skills but maintain/increase mental skills (if they have the capability). Younger players should be more raw physical talent until they learn the nuances of the game and develop their instincts (if they have the capability). Of course, have some randomness so more career arcs can happen and keep things interesting.
-Skills Training System. Much like the weekly prep of NFL 2K5 or the Practices of HC 09, players need a way to train off the field. This could impact week-to-week progression as well as the chances of potential increasing (or maybe decreasing depending on what you focus on during training and some areas get neglected, maybe borrow RTTS mode as well for this aspect)
-Stats could impact POTENTIAL not actual ratings. Why? Because scouts use performance to determine upside/project FUTURE performance. So a young guy that has a great rookie year is more likely to have his grades maintained/increased (especially if he surprised in an area). Likewise, a kid that disappoints could have the risk of potential lowering. However, if he has a great 2nd season, it could swing around, or he gets higher grades in other areas that might have emerged. A vet that's constantly performing year after year and maybe even getting better would maintain his grades and maybe even grow a few.
This could be governed by some sort of expectations system, based on draft slot for rookies/young players, and last 3 seasons average once a guy gets X years in the league.
However, the player would have to keep working hard to reach this potential and meet expectations that come with it - or else it might fall again. So the skills training will be what helps (or hurts) the player being able to get to/maintain his ratings, which determines if his stats have a good chance or not to meet expectations, etc.