I've been using YouTube front-ends for just over a year and it took some adjustments initially. There are a lot of moving parts to account for. My understanding is that many public instances (ie: yewtube) use upstream apps (like invidious) that use the YouTube API or yt-dlp themselves.
It's a constant cat and mouse game w/ Google. From my experience, if Google finds a way to block things, in a few days (or weeks) a workaround goes live and then works its way downstream, but it's up to each downstream client to install these updates in a timely manner. Then, on top of that, the standalone apps are still apps so they occasionally have completely separate bugs or changes themselves.
Personally, I self-host invidious through docker and then when it goes down, I check the issue section for more info and manually update my invidious when a change is rolled out. Separately, I have really enjoyed clipious (it integrates nicely with self-hosted invidious) and it still works for me, but there have been times when it went down on its own... for example, the developer turned on/off the proxy settings (can't remember which one) because ppl wanted higher quality streaming options and my app stopped working. Once I adjusted the setting, it was back to normal for me.
My suggestion, if you aren't doing it already is instead of switching apps, try switching instances and see if that makes a difference.