Hi @Jonathan,
Thanks for your patience here! I have a few notes from my colleague @SoilRotifer, which I'll summarize below:
It appears as though you have been playing around and trying many different things, rather than starting from a clean slate with your data. You should have merged your results after running all of the closed reference commands (different variable regions) prior to all of the other stuff. Also, there is no need for classification since you are using closed reference.
It looks like you have also been in a similar discussion with @jwdebelius on this thread regarding taxanomic classification on closed reference OTU clustering, where she mentioned the following:
One of the beauties of closed reference clustering is that the taxonomy is already assigned and the tree is already built for you. You just need to import them into QIIME 2 and you're ready to go!
After reading through that discussion, it sounds like there was some initial confusion as to why you were attempting taxonomic classification - unless you have any further questions regarding the motivation behind that, I would say that you should follow Justine's recommendations there!
Hope this helps,
Cheers 