Yes, with an exactly 50/50 chance system and zero trading cost or friction and enough trials, profit/loss will converge to break even. Has to be, or casinos would be out of business and every stock trader would be rich.
Setting aside the fact that none of the above are true in actual (as opposed to purely theoretical) markets... this does not mean that it's impossible to structure losing trades. The two are not mutually exclusive - there's zero reason for them to be - and I would have thought that this would be obvious to anyone. Extreme outliers come to mind immediately; so does chance of ruin due to incorrect trade sizing. There's quite a few others.
But someone arguing against the obvious, incontrovertible, fundamental, armor-plated fact that "reversing" bad decisions does not automatically produce good ones... that makes me wish for IQ filters. A "you have to be at least this smart to get on this ride" measuring stick at the entrance to this place. It's not just a bad take; it's a lack of simple cognitive ability. More than enough to qualify him for my "Ignore" bucket, anyway.