Somatic Hypermutation works on vaccine antigens as well, and diversifies B-cells so that the B-cells will recognize many spike mutations. Anyone who has been vaccinated or recovered probably has some B-cells which match this spike, even if they're not amplified and expressed into circulating neutralizing antibodies.

T-cells on the other hand don't diversify as much but they recognize primarily the host MHC section of the MHC-antigen complex, so they're already somewhat insensitive to mutations in the protein. The T-cells stimulated by vaccine or virus aren't special in that regard.

Arguably it doesn't really matter. If you get infected with this spike it is still going to have pretty much the same shape as the spike in the vaccine (after all even the mutated versions still need to bind to ACE2 and can't vary that broadly) and your vaccine-trained T-cells will still bind fine to the antigen-MHC complex that the spike forms.

There's this idea that antibodies and immune responses are like a key fitting perfectly into a lock where any pin being wrong doesn't unlock the lock, and that's just false. The immune system has a need to fuzzy match against antigens because it evolved against adversaries which mutate and attempt to reinfect.

You're pushing a narrative that spike-only vaccines are worse in the face of mutation, and there's just no evidence of that. On the other hand mRNA spike-only vaccines allow creating a very strong humoral response to the spike protein in the absence of significant side effects and their efficacy has been so far across the board superior to inactivated virus or viral-vector vaccines.

> You're pushing a narrative that spike-only vaccines are worse in the face of mutation, and there's just no evidence of that

"Pushing a narrative" is a pretty uncharitable interpretation of a short off-hand comment.

There actually is some evidence that the infected have a stronger immune protection to exposure than the vaccinated, and there are mechanistic reasons this would make sense, so I wouldn't go so far as there's no evidence for it. We'll just have to wait and see.

There's mechanistic reasons why the infected would have weaker immune protection because other viral proteins like ORF8 suppress MHC-I expression and T-cell responses. All the studies that I've seen that look at the problem properly find that recovery from infection provides highly variable protection with younger and less severe symptoms providing much less than vaccination.

>Arguably, traditional dead-virus vaccines might provide better long-term protection for this reason, but we went all-in on the new tech.

That argument sounds good in theory, but we have data we can look at to assess — Sinovac is inactivated virus and it's efficacy was 50% against the early strains. Also, work has shown that spike antigen vaccines broaden the range of antibodies produced by previously infected immune system to increase neutralisation against later variants, even though the vaccines were of course based on ancestral spike protein sequence..

There are only 3 proteins on the surface of the virus and spike is most prominent.

What is actually needed and is probably working for previous infection is knowledge about proteins that are generated after entering cells.

Antibodies and cell based immunity against these do not avoid infection but will allow early elimination of the infected cells.

This is again where mRNA vaccines can shine as they can introduce any protein onto cell surface to be recognized by the immune system.