This is a very interesting question.
I've come at ABA from an unusual route, which is via discovering PECS as a salt and thrn doing an MSc, so I know the theory and am still limited in exposure to grass roots ABA.
I have spent a large part of the last 4 years lobbying at managerial level for formla commitment to PECS which we now have in both health and Education. (Hooray!!!)
It is only now that I am able to watch and learn from people carrying out programmes and I myself have been mildly puzzled by the fact that some ABA therapists don't use pECS but focus on imitation tasks for speech.
When I studied EOs in depth I was staggered by the fact it seemed even necessary to do so. For a SALT this stuff is blindingly obvious. All disciplines are strong and weak in different areas but what SALTs are great at is emphasising functional -ie useful- communication irrespective of modality.
What we aren't good at (or able to do well given huge caseloads) is give precise and prescriptive advice on intervention. I realise that 99 % of people just can't take a suggestion or recommendation and go with it. They need stuff to be speeled out very clearly. This is what appealed to me about PECS and ABA.
I do think that in ABA there might osmetimes be an emphasis on repetition of speech with inadequate understanding, but that is why i thnk SALT and ABA can work so well together. They bring differenet but essentially complementary aims to the table.
In the case of your son DIL, it's impossible to pass an opinion based on what said here but if it was a RL professional situation in which i was involved, I would challenge vigorously and expect a very solid theoretical justification for the decision not to use PECS.