TLDR; Do I risk lingering issues (when continuing to final release and beyond) by upgrading to the 2019.3 beta rather than just waiting for the final release? E.g. like meta file and other junk causing problems?
Im saying production in quotes as this is just a hobby project (though I’ve spend a very long time on it).
I don’t mind the application being unstable and “buggy”, but I rather not mess up my project.
Well for me it is close enough before it goes into the final release that I already have my project switched over to beta due to that everything I have tested has gone really well and nothing stood out so mine is completely switched and I have no issues with mine. So I don’t see any reason why you would get issues either.
Thanks for the reply Binary.
Yes, I would assume most things being locked down by the time it hit beta. Its mainly the fear of “work in progress” features that, thought fixed within the engine on release, still possibly lingers within projects upgraded during the time frame.
I’m itching for the terrain holes feature 
In my experience you should upgrade only if necessary, otherwise you need to weight how much you will benefit from new features vs overhead work you will have to do related to project upgrade - which you really you can’t know until you upgrade 
I upgrade also when it is not only stable but also had at least 5-10 patch releases (eg 2019.3.5+).
But since you said it’s hobby project only and you dont mind some(possibly dozen hours) of additional work I wouldn’t worry too much.
Well the thing is, its a pretty straightforward (one click) upgrade for me that requires almost zero changes. Its a fps for android, and most of the stuff I use is legacy (i.e. no LWRP/Universal).
Its (was) mainly the fear of running into issues at some point into the future, directly attributed to having converted project with a beta version of the software at a prior date. That said, I would assume deleting and reconstructing the Library folder would clear things up 
Always back up your project before upgrading. The updates between betas and final releases primarily come down to bug fixes. Breaking changes are not expected. If the beta is working for you, it’s unlikely that upgrading to the final release or one of the following patch releases will break your project.
Thanks Leonhard, I’ll take the plunge.
Also at the same time, if something in the beta ISN’T working for you, reporting it before it goes gold will greatly increase your odds on getting a version that actually works on the release day
We tried moving our project to the 19.3 beta. We had to modify some stuff (using xxx.Experimental.HDRP to using xxx.HighDefinition for instance - shouldn’t that be covered by the automatic upgrade process btw?) but it was quite smooth.
We had to modify some code to make it compatible with the new play mode options. Totally worth it.
However, we didn’t switch in the end because of this issue that breaks most of our mesh colliders : Upgrading to 2019.3 results in all MeshColliders requiring readable on mesh