Evolutionary pressure selects for less deadly viruses over time. You're right though, there is still longitudinal risk as this does not preclude a vastly more deadly strain--but at that point the effectiveness of a vaccine is similarly unknown.
The virus has little evolutionary pressure to be less lethal. COVID has already spread before the symptoms appear.
I'm hearing that for the first time, do you have a source?
> longitudinal risk
With longitudinal I meant the risk of what we have labelled 'long Covid' but really seems to be a conglomerate of different, sometimes chronic diseases in different people, that we still don't really know how to treat effectively (not very frequent in absolute terms, but orders of magnitude more frequent than any serious side effects we have observed due to the vaccines).
So if I understand you correctly, what you are fearing is that there is some unknown risk about the mRNA vaccine, that will only reveal itself over time, but that we have no idea about the mechanism of action nor any actual cases. Is that right?
> I'm hearing that for the first time, do you have a source?
A virus that kills its host before it has a chance to infect others cannot reproduce.
I cannot remember which virus it was (might have been ebola) but it wiped whole communities so fast it died off because there were no new hosts to infect.
I also want to remark that there is the opposite effect that transmissibility correlates strongly with viral load (amount of viral particles) and viral load correlates weakly with lethality.