I was using the same language as the OP when referring to the poll, but if you feel better about yourself now good job.
You draw a pretty extreme conclusion about the polling of a generic candidate. Honestly it sounds like another specious conclusion that's been drawn because it agrees with a bias. I'm open to being wring and am interested in how you came up with it.
This CNN article has some pretty interesting discussion about generic candidates. The general consensus seems to be that generic candidates simply indicate a party preference rather than a judgment about a particular candidate.
I think I see what you're trying to say. You're saying "the best candidates are the unknown people, the nobodies, because Trump is getting 46% to 47% of the vote against them, rather than the 48% he's getting against Biden".
I drew the specious conclusion that you were refering to Joe Biden as the best candidate because he is polling the highest among candidates (tied with Harris) at 45%, has nearly the same margin of victory against Trump as all other candidates (2%-3%), has beaten Trump already, already has a massive campaign infrastructure, and is the current nominee.
On your last comment, more important than Donald Trump losing 2% to "Not Sure" is the fact that he's still beating all the candidates by 2%-3%. Without more information the best assumption we can make is that the undecided voters will vote the same as the decided voters once they have enough information.
As I said before, the only real conclusions we can draw with certainty from these polls is fewer people know who these candidates are than know Joe Biden.