I think the best solution in this situation would be Core to support Segwit2MB.

Do core devs say that there are many desired improvements that need to be packaged together to do a HF once instead of several times? I think most people would understand that. Ask for reasonable amount of additional time to do that. Cannot be It's not negotiable.

Is 2MB base size slightly more than Bitcoin needs right now? Maybe so. But should adoption grow, Bitcoin will need it anyway.

Are we afraid of 8MB total size blocks? I don't think so. Even with transactions backlog like now, most of time we won't be having 8MB blocks, more likely most of blocks will be about 4MB. Even if increased storage, bandwidth requirements will force some people to give up their full nodes, small decline in full nodes count isn't critical. Moreover, increased adoption may result in increased number of people interested in running full nodes.

So what's the point in refusing the 2MB HF and pushing the UASF? To teach miners a lesson? Maybe they deserve it. But it's unwise, immature. Even if they are so, does that mean that we should be like them?

If Core accepts the 2MB HF (not just plain blocksize increase, but combined with all desired improvements, properly scheduled), Segwit will most likely be activated sooner than with BIP148, in more reliable, civil way.