A few comments, though not on the underlying immunology
The good things about this study are the huge D-IBS sample (n=2,375) and the focus on the comparison with a disease group - Inflammatory Bowel Disease, rather than healthy controls. That might make it more clincially useful.
On the other hand, the results v healthy controls were very unimpressive, and I'm always suspicious of a study that's been hyped in the media - not to mention the immediate marketiing of IBScheck (TM) as a diagnostic test. That's premature, in my view. The biggest flaw (common in these 'biomarker' studies) is the failure to replicate in an independent sample, so the accuracy of the test is cherry-picked, using thresholds set to maximise accuracy in this particular set of data. The D-IBS sample was big enough to split into test and validation samples, which is how you are supposed to do these kind of studies. Peer reviews asleep on the statistical treatment, it seems.
Even with these cherry-picked thersholds, the results aren't that good, with most cases of D-IBS missed (the best they achieved was a 60% successs rate - barely better than chance - but using that particular threshold meant a lot of IBD (disease) cases were misdiagnosed as D-IBS).
I still think the study is interesting and worth following up, but they hype and rush to market are very disappointing. The ScienceDaily piece was merely churnalism, reworking the IBScheck press release.