Are you going to try to talk with the professors personally? How and when are you going to do that?
edit: furthermore, how do you know the professor can single-handedly have any impact on who gets admitted into the program?
Sure, that's a pretty good idea.
> How and when are you going to do that?
Figure it out. If you want a cookbook solution, the answer is probably to just play the GPA game.
Even if you don't get the position, there's still a lot of value to be had from a conversation (for both parties).
My perspective may be too cynical though.
When a professor writes that on his/her personal/research website, it means "I already have grad students coming out of my ears, and I don't have enough grant money to support even half of them." This is useful information, because you (should) pick a grad school based on who you want to work with. You need to look elsewhere if all the professors you want to work with at a particular school are over-subscribed. Some professors also become jaded by the sheer number of unqualified candidates who can't hack it: None of the string theory groups in my school will talk to you until after you've been admitted, gotten good grades for a couple semesters and the passed department's second-year screening exam with a good score.
Let me part with this: I got into grad school (probably) mostly because I sent an email to the professor who ended up being my research advisor. I described what I did in the past (which was sorta-kinda tangentially in the same field), omitted any mention of my (not very good) grades, and asked a non-time-waste-y question about the research group. The thing about grad school (at least in the hard sciences) is that one of the criteria for admission is "doesn't anybody want this candidate in his/her research group??" When you have someone pulling for you on the inside, getting in is a whole lot easier.