The first thing that comes to my mind is that playing just one table will make it much much much more difficult to evaluate the long-run effectiveness of your play.
Secondly, rake at 25NL and below is so massive that getting involved in a large number of marginal situations will likely decrease your winrate. For an illustration, if you make a 55:45 flip with no rake, in Sklansky bucks you win 55/100-45/100=0,1 times the amount wagered. With a rake of 5%, you win 55*0,95/100-45/100=0,0725 times the amount wagered. Here a 5% rake will make a roughly 30% difference in the amount won. If pots are capped regularly it is effectively a smaller rake, so with a cap a significantly larger number of marginal situations remain profitable. For an opposite example, if you make a 100:0 flip, then 5% is exactly the amount of profit you lose.
I'm just learning as well (making ~5ptbb/100 at 10NL 8-tabling as it is), but I definitely think if profit in upper microstakes internet games is a worthy goal for you, then in full ring you should on average be playing at least 4-6 tables even when learning, and instead of loosening your range in a limit you beat you should in general rather attempt to move up with a tight strategy. If you play 14/12, money doesn't just gush out of you even if you're playing geniuses. It's a small dribble at most. If you attempt to take your already non-optimal 25/15 to a place where it doesn't work at all, you'll bleed to death quickly.
It's definitely a learning tool that can give you ideas, but your main direction should probably be taking TAG ABC as far in limits as you can, THEN loosening up there a little until you can move up again, THEN loosening up there a little etc.
That being said, I'm definitely playing less than optimal for learning purposes
Actually, I think you should do that a little
But I'm talking about 15/13 instead of 12/10. 25/15 is too crazy.
BTW I still doubt 25/15 is ever better than 25/20, whatever funky situation or purpose. Limping just isn't cool.