Agree entirely. 2.2.x is just the dedicated "We are primarily fixing 2.2 stuff, but also other stuff when we notice and have the time". 2.3.x will just be "We are primarily fixing 2.3.x stuff, but also other stuff when we notice and have the time".
It's not like after 2.2.6 Paradox completely shuts down the bug reporting section or something.
I think the concern is not that Paradox will *stop* fixing bugs, but that they will go back to
adding them faster than they remove them. Even 2.2.6 - an update that followed a lengthy 2.2.5 beta, and that was well into the 2.2.x release period - added some major bugs. Some of those have now been hotfixed out, and hopefully the non-beta release of 2.2.6 will squash the rest, but it doesn't inspire confidence to see an obvious
new bug (the planetary stability events issue, which comes with pop-ups and everything) appear in what was supposed to be largely an update about fixing things.
Meanwhile, there's still a bunch of significant 2.2.x bugs that have never gotten fixed. For example, it's still possible to insta-repair all buildings on a planet by sacrificing one building that is to the left of (or above) the damaged ones, and wormholes can still be pathed through without exploring them properly. There's also more 2.2.6-specific regressions, such as a weird interaction between updating and ship design, where I was unable to save
completely new designs with unique names while any of my fleets were updating.
I am absolutely 100% not one of those people who decry any suggestion of a new feature with a wail about fixing bugs / improving AI instead, but I do want to see Paradox put a lot more effort into QA. While 2.2.6 was arguably more than just a bugfix patch - we got a UI overhaul, a significant AI update in an area that seems mostly outside of what Glavius or other modders do, and a major QoL change to ship updating (thank you *so* much for that!) in addition to the usual slew of balance and bug fixes - the initial state of 2.2.6 didn't bode well for any update with more dramatic changes. I worry that 2.3.0, minor though it might be, will be another 2.2.x (where I desperately hope x doesn't equal zero) and we'll have to wait months again while (some of) the kinks are ironed out (and others added).
Getting 2.2.x into an actually really solid state - rather than the "this is technically release quality but you could do a lot better" that we have today - would probably help a lot before they start messing with 2.3. Have a firm foundation and check for regressions as new code is added, rather than just checking the new code (though please, seriously, do that too!). I realize this is unrealistic and again, I don't want to be one of the "no features until there are no bugs" folks, but I am still deeply concerned by the idea that, because 2.2.x has finally reached a decent state, the current system work. History shows that they'll have the hands more than overflowing with bugs once they start work on even the smallest non-cosmetic update.