I've given myself a deadline to resolve all my issues and release the first official version by the end of September. If I can't meet the deadline, I will step down from leading the project and transfer full rights over the repository and instance to the contributors.
Ernest, please don't be so hard on yourself. Deadlines slip, even for products formally released by companies, and this is more of a hobby frankly. I think what might help is less of a deadline and more of a roadmap - like, here are the major bullet point items we want to target for release by end of 2023, by end of Q1 2024, and sometimes those slip but then the roadmap can be revised.
I've been updating my own kbin instance pretty regularly, every couple of weeks, and I've seen things become more stable over time (less frustrations in upgrading, more features, etc). I'm quite happy with the progress so far. This project has grown so much in such a short time, and the fact that the kbin issues matrix is much quieter than it was speaks to the growing stability of the platform.
As far as kbin.social itself, I would agree with some other folks that you might need more volunteers on the actual instance administration and moderation front.
And as far as spam - email, the original federated messaging platform, still has that problem! Each email provider has to handle it on their own, using increasingly sophisticated methods, and they're still not perfect and it's been decades. Yes, spam is frustrating, but due to the nature of ActivityPub we will always be in an escalating war with spam. It will never be solved, only mitigated for a time.
Anyway, perhaps I've written too much here, but I have a ton of confidence in this project and also in you, and I hope you look back and see how much has been accomplished in a short amount of time, how much kbin.social has grown, and how the amount of other contributors indicates an overall great level of confidence in what you've created.