The stats are a bit meaningless because Nc6 and e6 are interchangeable and also, they lead to at least four different plans. The classical, where black plays e6, Nc6, d6, Be7, Nf6, e5 etc. The "crazy" where black plays e6, Nc6, Qc7, Qc7, Nf6 and tries to mate white, the Qc7, where black plays e6 and a6 but delays Nc6 and of course plays Qc7, and of course the really dangerous variation for white.
the STATS tell the TRUTH about what happens. google cognitive dissonance
math lesson: 52% MINUS 44% = 8% (MORE THAN 44%)
that's white WINNING 8% MORE of the time than black, and the flaw with YOUR argument is this isn't bizarro world
OH! WAIT! maybe it IS!
of course I'm POSITIVE there ARE LINES where black can do better, but statistically THOSE are meaningless "negligibilities" which I just neologized
the FACT is, there isn't a single line that's a refutation all the way up to 2500 where white is winning games by a minimum of 1% in the main line. rare stockfish generated lines that 20 people know aren't refutations... they're boogeymen
looking into the paulsen main line, black CAN get a couple percent advantage in the main line IF white co-operates with 1. e4 c5 2. d4 cxd4 3. c3 dxc3 4. Nxc3 Nc6 5. Nf3 e6 6. Bc4 a6 7. O-O Nf6, but doing one's homework and simply playing 6.e5 instead puts white right back into 52:44 stats, and I'm STILL in 2200-2500!!! at the 1600-2000 range, 6.Bc4 is totally the way to go for juicy 51:45 stats.
on what planet is an 8% edge "nearly losing"?
black ain't winning doodoo in ANY line if the gambit's accepted
even filtering 2200-2500, black is STILL NOT winning in ANY LINE, so you either live in a fantasy world inside your head, or are a hater and a liar trying to steer others away from your nightmare