Thanks so much for the files!
Have you had a chance to check out the community plugin, q2-fragment-insertion? This community plugin wraps sepp, which should help out with what you are working on.
I was perusing the source code for this community plugin and noticed the following comments:
Beta-diversity computation with Qiime2 requires every branch to have a length, which is not necessarily true for SEPP produced insertion trees. Thus we add zero branch length information for branches without an explicit length.
I think that means my earlier hunch was correct about not all branches having a length, but fortunately, the q2-fragment-insertion plugin looks like it can do the cleanup for you to make sure this phylogenetic tree is compatible with downstream QIIME 2 analyses. Keep us posted! 