I'm not "pushing people to LN". I'm currently not even using it actively myself (after some experiments in 2018 or 2019).
But I would use it if the fee situation of late April, or some months in 2023, became the standard, at least for some transactions.
For the "becoming more centralized" assumption, yes some big blockers in this forum continue to repeat that, but that's pure speculation. I haven't seen really research on that, if you know some, then please point me to it.

"Network support" is not the only benefit a full node provides.
Their owners may use them for better privacy and more safety against some attack vectors possible with SPV clients (not that this is something an average user has to fear, but somebody doing regularly big transactions could)
Using a client like Electrum you leak a lot of information actually if you don't create lots of wallets and do a high amount of address management. So the option to use Bitcoin Core with an average PC should imo not be sacrificed.

But isn't that just supporting what I wrote? Monero nodes are much less hardware consuming than Bitcoin's currently. That's mainly because the network usage is lower. But I think that's the issue you're discussing with BlackHatCoiner ...

I think the Monero conditions cannot be simply assumed for Bitcoin. Bitcoin has much higher usage. Thus, to have the same fees you would need huge blocks.
I had in the past advocated for a dynamic but still quite small block size. But I think now the "L2 way" is more promising, because doubling or tripling block size, which would be ok still for today's consumer hardware, would not be enough really for mass adoption.
About tail emission: I already wrote it's an acceptable model for me. But the problem is that Bitcoin's value proposition currently draws a lot from the "limited supply" and "digital gold" narrative. In addition, I would prefer to not hardfork forever, even if altcoins are accustomed to that. If Bitcoin was created newly I would advocate actually for a tail emission model.
And Bitcoin can work with transaction fees only, that's the conclusion I get from all current statistics on the subject I know. For coins like LTC I guess the situation could become more critical, but LTC miners actually benefit from Doge's tail emission.

Myself I think the "multiple L2" model, with sidechains *and* LN (not *or*!) could bring actually much more benefit, and it would not depend on tail emissions.

But if the security situation really becomes critical, one could implement something very similar to tail emissions actually with Drivechains or other kinds of merge-mined sidechains, with a model very close to the Doge/LTC model I showed above. If each Drivechain has an own utility token which is inflationary, then the Bitcoin miners would actually benefit from merge-mining them. For this reason, if Drivechain or another decentralized sidechain model is implemented, I expect the Bitcoin miners benefitting quite well from them.

Regarding miner centralization I have already written in another post that I disagree that there's a clear tendency towards centralization. I could also speculate and say it's actually the other way around: big miners aren't flexible enough for the rough post-halving conditions. Just a few days ago I saw a new thread in the German subforum of people (real forum users, not forum users talking about something they heard "from friends"...) setting up small solar mining installations.

A heavy crash due to the "deflation obsessed investors exodus" could also go so deep that it would hurt security itself and cause a lot of chaos, maybe even it's the opportunity to a anti-Bitcoin attacker.

Thus my opinion continues to be: We should try to preserve the current Bitcoin mining model and, above all, avoid hardforks. If really (for me, likelihood is low, I'm not "praying" for it) something goes wrong and security becomes too low, then we have some time for alternative solutions like merge-mined Drivechains, and only if that doesn't work we should consider a hard fork to tail emission.