And I’m okay with your example, that would be to me best described as “even if it was for reasons you would like (defeating Trump) he did actions you think are bad (neo-Nazi-like, and specifically transphobic)”. Like you can say that Newsom shouldn’t belong in the Democrat tent because of this or that, but if he proposes housing policy that you think would be helpful, either link a material reason as to why his transphobia, previous deed or other negative quality would taint this proposal. Otherwise say something “I like this policy …even if I don’t like him” or “…even if he’s a shitbird neo-lib transphobe” or “…even if he’s probably just doing it to run for a future Presidency” or “…though most of the credit should go to the CA Assembly”.

In a more extreme hypothetical, if Trump were to somehow get Grok or ChatGPT to slop out a universal US healthcare policy document that has comprehensive detail, I might applaud the plan itself on its merits, but of course I know Trump is a pathological liar, changes his mind all the time, his administration is full of idiots too evil and incompetent to implement it, and Republican, big pharma and insurance donors will never let that get off the ground and so I’d have little trust in that happening. But I would say “Trump, as much as I despise him, had a good idea for once that Democrats could actually try implementing for real”, or “he’s probably going to say the opposite after a quick chat with Perdue” instead of “I don’t like this plan only because it came from Trump”. Discourse would be better if we could separate the words/actions from the speaker, at least to start, but say why that speaker or a relevant larger context makes the words/actions unreliable if that’s the case.