The main point in releasing only once it is
good is that it is released when it is perfect, so you don't get e.g. desperate bug reports, or complains about unfinished areas, or snick comments about stuff which you had to cut corners. Plus, you can rework renders and none will complain if you spend a month reworking stuff rather than adding more content, giving you a lazier workflow and
hopefully a healthier working environment.
The downside is that occasionally people complain about the time between releases, so it relies a lot on trust.
And it should go without saying that this is a proven method... for yearly releases, or stretching, two years. At some point, this method will cause you
more mental strain than actually releasing unfinished stuff. It will create too many expectations which you'll not be able to meet, and to actually meet them will cause you to use so much time that it'll become an endless loop.
But someone once said, "there's always time to make things perfect after the 1.0 release". For example, if you start learning AI-generated imagery, you could reach a new level. Master ControlNet, and you will surpass almost every other game which exists here. Do this now, and....
...you'll never finish.
And by not finishing, you'll drain more and more the trust and eventually you'll be left alone with haters. This will demolish your mental health. You should avoid this situation for your own sake. Instead of seeking quality and keeping those whom trusted you happy, try organizing your own schedule. You can't make anyone happy if you never release, not even you.
TL;DR In case Komisari failed to notice, they passed the point where making a quality release causes more satisfaction and less mental damage than releasing buggy stuff often. I advise getting help, even if they end up hiring a project manager instead of "extra hands". Keep in mind and take advantage that Patreon pages can pledge to other pages.
But most importantly, Patrons are feeding you not only money, but trust. If you can't trust your patrons enough to eg. show them all the ugly hacks you need to get stuff done, or how many liters of coffee each render costs you, or demos with a longer list of "known issues" than actual "changelogs"... Yes, trust was never meant to be uni-directional.
Good luck!