Would you play differently at 100% anonymous tables?

The reason I'm asking this question is because I watched a video where BenaBadBeat and Uri Peleg discuss hand histories.

https://www.youtube.com/watch?v=2m9Ht4Mw...

At 3:00, Ben says his two sizings on this board are b33 and pot, but his hand is too thin, so he checks.

However, Uri suggests b10. But Ben says it would be impossible to balance his b10 range with stronger hands.

Then they go into a philosophical discussion: Uri says "How is your opponent going to know whether your range is protected?" It would require a massive sample size to know whether your specific opponent is balanced or not in a unique river spot.

So, my question is: When you have a "reputation" based on thousands of hand histories, perhaps it makes sense to be more balanced and thus less exploitable.

But if you play at anonymous tables without a "reputation," then you can't be exploited based on your actual hand histories. You can only be exploited based on population tendencies. So, in that case, it's okay to play "unbalanced" as long as your opponent doesn't believe population is also unbalanced in this spot, and will correctly adjust?

I'm curious what people think.