Since the spike protein has to bind to human receptors, it has an optimal shape.
It also binds to antibodies, which resemble the receptor.
It is like a bad copy of a key, that works in 2 very similar locks.
One lock is the receptor, the other is the antibody. The more mutations that pile on, the greater the chances of failure, not success. If the conformation made by 3 mutations fits our receptor better, the chances are
it will fit the vaccine's antibodies better too.
Getting a version that does one better but the other worse is hard, by random chance. This is what we get by chance:
What I'm saying is if the spike escapes the antibodies, it
will likely also escape the receptor- and that variant is a dud that will just die out. Like any random one up there, that can't fit either lock - It will lose the competition.
Until actual data shows that vaccines are losing efficacy, there is no need to panic or assume bad things.