They're probably saying that because scientific conclusions usually err on the side of caution. Plus they're using anecdotal evidence to make their claim which is known to be unreliable. However, there's really no other way to gather data about this except asking people or maybe sticking them in an MRI machine while they achieve orgasm which would be awkward and expensive, so it's kind of the best we've got.

They can then compare cisgender orgasm descriptions to transgender orgasm descriptions and see if they match up. And despite the shaky foundation of anecdotal evidence, this is a very long term study and they used a pretty large sample size (130 participants) so it should weed out any anomalies.

It's certainly still possible that the participants are imagining or embellishing the results, but I think the people behind the study did the best they could to gather accurate data.