I agree with the spirit of your comment in that I would hold out for more scientific replication of these results to increase my confidence beyond the level of simple curiosity.
Traditional vaccines at least dump most of the target viral proteins into the body.
Focusing on a single protein that may mutate in the target virus has always seemed myopic to me, let alone the assumption that this approach could possibly be more effective than an immune response to the actual virus. I’m not saying that this approach couldn’t work, just that it would, by definition, trigger a subset of the immune response of a whole virus vaccine.
Edit to add: by better, I mean better at triggering an immune response. I’m not arguing against the idea that a targeted, single protein approach is or is not better at reducing illness or death and therefore shouldn’t be pursued, just about the idea that recovered individuals would be less protected than a vaccinated individual on average.
I don't think its so much about the idealized immune response, but that in general, people's immune response to natural infection seems to widely vary, with many not having much of a robust response.
It very well may be that a good immune response to natural infection may be better than one from an mRNA vaccine targeting the spike protein - but in aggregate the vaccine will be more successful because its induces a more consistent response.
Is there academic research on this topic that existed prior to SARS-Cov-2?