Usually in a particular opening position, if move A is more popular (in GM games) than move B, then either move A is objectively better than move B, or move A is equally good as move B.
Are there any examples of moves that move A is more popular than move B in GM games but move B is objectively better?
It really depends on how broadly you want to define the openings.
For example, while not popular at that level, GMs have played the Smith Morra. Of course, a standard open Sicilian is better.
The same goes for those, who have played the King's Gambit when trying for the Ruy Lopez is better.
The Ruy Lopez is probably objectively better than the Italian Game, but we saw the Italian played the last two days at times in the World Rapid and Blitz events.
The same goes for the Benoni, Benko, and Dutch. They are not horrible, but there is a reason the do not dominate high-level opening play.
Anytime a GM needs to win, and plays a sharp opening that is not objectively best, and that happens a lot.
I am sure there are plenty of examples that are deeper into the opening, but this already shows how common it is.
Lastly, many GMs play sidelines, and have pet openings that are not considered best. They play them when they need an imbalance, and to win.
In general, no. The games between players of higher levels (going from beginner --> amateur --> master --> grandmaster --> correspondence) tend to be better, and they are what usually dictate theoretical trends. If there was some well known line that was dubious, a GM couldn't get away with playing it often because their GM opponent would be prepared.
There could be a dubious line that was more popular at higher levels, but in this case the sample size would most likely be small (meaning the move is a surprise weapon). If many GM games featured this dubious line, it wouldn't be a surprise weapon anymore.
Playing good openings can be thought of as a necessary condition for reaching the highest levels of chess. If you consistently start the game poorly, you most likely won't reach those levels in the first place (or won't be able to maintain your status there). There may be some rare GMs who can get away with this against their contemporaries, but such a style of playing would never work in correspondence (the de facto highest level of play).
