lol... you're kidding us, aren't you? It's an abstract. The...

lol... you're kidding us, aren't you? It's an abstract. The article will contain a full, dry, detailed description of:
o study design and protocol, description of the tests conducted
o raw numbers
o statistical analyses
o conclusions

"results were significant but against what" – the control group! This will take the form of a level of confidence that the difference between the active and sham arms is not due to randomness. To be regarded as "significant" the confidence will have to be greater than 95%.

"Autism has so many variations which cannot be treated all the same" – But there are standardised tests that are used to diagnose the degree and placement of the subject on the autism spectrum. As noted in the abstract subjects were measured using these standardised tests before and after being treated with the device. The difference between the two measurements is recorded against whether the subject received active or sham treatment. You'll have to wait for the article to see what the differences were, and what the confidence level is.

"Id say that they are going to release the results which dont say a lot and then do another CR to ramp up production" – You're joking again, aren't you? Frontiers would absolutely not publish an article that has no substance. Competition between journals for primary publications is intense and ferocious! The reputation damage from publishing a marketing piece as serious work would take decades to repair.