Thats the issue with balancing though. What works in high level play doesn't necessarily have to work in low level play.
So it becomes a question of for which level of play do you balance. I'd argue that whatever meta happens in higher levels of play will slowly "trickle down" to all levels of play, but it takes time.
A good example for high vs low lvl of play is the light class.
The light class sucks major ass in high level of play where teams are coordinated and everyone has good enough aim to hit a L and simply win by HP diff.
In low levels of play the L class can be very oppressive since players aren't very coordinated, also tend to play the L class (instead of only playing M or H) and cant generally aim as well so the L class does have some windows to kill players, especially out of position players.
Over time, even low skilled players will stop playing L though, because there is a very real reason why L isn't played in high level. These people are called META slaves. They blindly (but for a good reason) follow the meta, don't even utilize the actual reasons behind why stuff is META, but simply by blindly playing meta, they will on average have higher chances of success.
What I would argue for needs to happen balance wise is to definitely look at success rates of each weapon and class across most ranks (gold-diamond) or something like that. Success rates being which class/weapon makes it to the final round and how often. I am fairly certain from personal experience as well as other peoples experience (streams, youtube) that the L class is the rarest class to make it to the final round by far. Across all ranks. Further evidence is the LFG. Literally NOBODY actively looks for L teammates. Nobody stacks L as well. L is just a really bad class compared to M and H, simply because of the HP diff. If all things are equal, an M and an H always wins. Simple as. L would need to be dealing more than 2x the damage of a Heavy for the HP diff to even out.