According to NIP-51, it should be possible, but it would be up to clients to implement it, and it's not 100% obvious how follow and hashtag block would interact.

Having thought more about this, I think clients that implement Mute lists from NIP-51, handle hashtag muting, and also implement NIP-02 contact lists (pretty much all of them) should provide either a switch to determine which takes priority, or clearly document which takes priority.

A potential enhancement to the protocol would be to add a "Prioritized Mute" and "Prioritized Contact" list types, that include a numeric priority, and allow them to be shuffled together in a way that makes sense. But that's really complicated from both a client and a user's perspective, and has some potential confusing side-effects while we wait for all clients to support it.

Another possibility I thought of while typing this up is for a client to show all content for any followed user or hashtag, but if it contains a hashtag that's blocked, treat it as if the content-warning tag was applied (which I guess warns you or blurs it out rather than show it by default?) I don't actually know how, or how many, clients implement the content warning functionality. I may see if I can test it.

Now that I know it is not currently possible, I can stop looking for this function.