The version A) is a big issue actually, thanks to the Hoskins effect.

That means that if the immune system is trained for a slightly wrong type of pathogen, it might have a worse immune response to the actual pathogen at hand than if it wasn't trained at all.

So if that viral protein would mutate a bit, so that it's still recognized as viral protein by the immune system, it might cause the immune response to the actual virus to be worse instead of improving it.