Otherwise we risk getting into a "king in the high castle" situation where the developers do primarily what they want, and they very rarely try to communicate with normal users.
It is double-edged sword: if there are many different clients, then some bug may be present in one client, but not the other. And then, there is a risk of forking, if some mining pools will run one version, and different mining pools will run something else.
Yes, both situations have their own respective risks.
It is important to stick to a single version during early stages, but now, I guess the environment is mature enough, that we can work with multiple clients.
Perhaps, but some of the demand for it would decrease if Core developers start behaving better. It is a bit ironic at times. They want as many people to run nodes as possible, but when these same node runners have concerns about stuff like OP RETURN then the developers become quite dismissive of them.
But in general, Bitcoin Knots is very similar to Bitcoin Core. There are some minor changes here and there, but it is not something rewritten from scratch, and implemented in a completely different way, than Bitcoin Core. Which means, that many fixes, and code changes, are used by both clients, and they share most of the source code.
It is a good setup, as the risk of forking between the two clients is very low.
If I recall correctly once he even advocated for lowering the block size.
If transaction batching would be widely deployed, then it could be done. And also, signet blocks were already reduced from 4 MB witness into 1 MB witness, just to test smaller blocks, and build fee pressure, without the need to artificially make more transactions.
Also, fortunately, block size decrease can be done without any fork, all that is needed, is just convincing enough mining pools to do that.
I don't think trying to artificially force fees up by lowering capacity will ever work. If anything, we need more capacity and minimum fees. It would be a better solution, but we should continue this in another thread if we want to talk about this.