One complicating factor with a lot of these studies is how "autism symptoms" and the like are measured. Specifically, they measure something like "how much distress is the autistic individual under compared to how much they can handle". High level of distress leads to meltdowns, inability to handle outside behavioral demands, etc, and shows up as higher autism symptom load.

So this study is, IMO, much less about "gut problems cause autism", and I highly doubt that these individuals get any less autistic. A fairer characterization is "autistic people with gastrointestinal distress are less able to deal with things in general, and fixing the GI distress allows them to handle social demands better".

That depends on what Autism is/how we define it.

I think there may be people whose brain works like an autistic brain in that the things that specifically distress autists also distress them. But they are good enough at tolerating distress that they don't actually show the symptoms of autism.

If people like that exist and we don't consider them autistic, then improving someone's ability to deal with distress can indeed reduce or even cure autism.

While the gut is one factor for some autism it is not the only factor.