We used to point out games not working well natively on Linux for what they are: bad ports.

We still do.

We used to ask for ports for each and every game we wanted to buy.

... which basically never worked for obvious reasons.
Even now, with a share quadruple of what it was it still wouldn't.

Devs either do a native port for intrinsic reasons or they don't - outside pressure from an absolute minority is a complete futility.

Nowadays not only "I can run it." seems to be enough for most people.

I am going to assume the first "not only" is too much, otherwise I don't get that sentence.
If you can run your game without issues, you can run your game without issues. If the performance is good, it is good.
Everything else is secondary at best.

They even started to actively ask developers to not port games.

If a dev already shows clear signs of not being up to the task of doing and supporting a proper native port, that's an extremely good suggestion.
We've had plenty low effort, zero support ports.
And we had plenty good intentions of devs, who then realized there is actually some work to supporting an additional platforms, and as a consequence stopped. I'd argue it objectively would have been better if those never even tried, because their failure story is guaranteed to carry more weight than any success story (negativity bias & media echo).

Devs already comfortable with Linux are very welcome to do a native port.
But those who never even touched Linux, have no real intention to, and only push the "export to Linux" button? That's a huge risk likely to come back to haunt Linux gaming as soon as a platform-specific bug arises.

They even started to attack fellow Linux games asking for ports.

Yeah, I doubt that - at least no more attacks than anywhere else on social media for any given topic or opinion.
People are people, the social climate is what we live in, it is what it is.

Using Proton will always be being second class citizen.

A clear improvement over what Linux used to be: no citizen at all, and third class or worse often enough even if there was a port.

Proton will always have to chase whatever Windows comes up with.
Proton can at most be on par, but never ahead.

Nice platitudes. But:

What more than "on par" do you need when it comes to video games?
What do you actually want to be ahead of?
Run them faster? Video games are mostly platform agnostic code, any kind of serious performance bottleneck will be cross-platform. At the bottom of it all is usually a C/C++ (or other low-level language) core that is standardized and will thus also behave the same no matter where. MSVC, gcc, Clang, etc all perform within 5-10% of each other, with differences only showing up in (surprise!) platform-specific code.

Features, like eg certain GPU techniques? May be thinkable, but that, too, is mostly hardware-bound.
But nobody would develop such a thing Linux-first, with a 2-3% share.

The gist of everything is that for anything serious to happen, the share has to rise.
That "No Linux support, no buy" crap didn't lead anywhere. Neither did the friendlier variant, the "Linux port pls" begging.
As Valve and years of dedicated work on compatibility layers has shown, you rise the share not by pestering devs with minority demands, but by ease of adaptation and support. And good hardware ideas.

But how about letting others have it their way?

Nobody is stopping you.
You'll just have to live with most people in the Linux community having realized by now that blind idealism leads nowhere and consequently calling others out when they show just that.
Somebody telling you to "just use Proton" is not an attack.