#59 There's a difference between a player being given an opportunity to win and squandering it, versus being stripped of that opportunity. I have never seen a game where human players played the tablebase moves resulting in a draw in an otherwise won position.
Just because a player hasn't won K+B/N versus K+any doesn't mean they should be deprived of their 50-move opportunity to do so.
There are many completely winning positions where a clearly winning player has gone out of their way to selfmate or lose over a lengthy series of moves. I've lost OTB tournament games in this fashion, and many opponents return the favor; but my own games are insignificant. Although generally K+B/N versus K+P is only lost if the opponent underpromotes, the same can't be said of any other K+B/N versus K+P+any.
And Sarg0n is right, if you start to define single cases as drawn you open an immense can of worms; USCF has been selling rulebooks for decades and with each revision still fails to address problems such as in #12 (dead position) or #60 (mate in 1 denied).
The whole point of USCF's ILC rule (as opposed to their "sufficient mating material" rule for claiming a time win) is that it's supposed to apply to positions where the result is self-evident, for example K+P versus a full army, with the pawn blockaded and mate imminent should not be a win for the K+P. Ditto for K+P (a-pawn) with the opposing king blockading the pawn, or K+R versus K+R (in most positions), etc. which are far more likely to occur in time pressure (and be unable to liquify into a forced draw) than K+B/N versus K+P is. How can K+P versus a full army, with the pawn blockaded and mate imminent (or similar positions covered by USCF's ILC rule) be declared lost (where a draw or better is self-evident), but positions where K+B/N has winning chances be declared drawn (even including positions where it is likely for a player to fall into a mate, or a forced mate exists)?