Only if you have a set of bacteria, call it pathogen Pi, i= 1... n that are similar except that they differ with property i AND that the organism that is infected always are infected by one or more pathogens of type Pi. Then if you take away a specific pathogen with property k say, either by vaccination or a very specific antibiotic that kills all “k” then the k-resistant strains (= all others) will take their place, if they can survive the competition from other bacteria, and not only of type Pi.
But, this is not how it works, you are not always infected. So the objection is not valid. Hoverer it is valid if you have a test tube and are growing all types of Pi, then if you eliminate a certain Pk its place will be taken by the other Pi with i not equal to k.
You may not always be infected with a resistant strain.
Someone else will be.
How do you think we got MRSA?
The current pertussis vaccine escape mutant has a mortality/morbity about 10X greater than the original strain. And the vaccine has zero effectiveness against it.
Try for a 100X strain?
But wait!, there’s more!
http://www.scientificamerican.com/article/baboon-study-reveals-new-shortcoming-of-pertussis-vaccine/