Looking too closely has zero upside and plenty of downside.
If you’re wrong and the result was legitimate, you’ve developed a reputation as a pain in the ass.
If you’re right, the researcher will say it was an honest mistake (maybe a clerical error) to absolve themselves modulo retraction, which happens all the time. Nothing of significance comes of it, but an “influencer” may well have other, above board, programs of research on which you could have collaborated and made a real name for yourselves.
Meta has been aware of DeepSeek for a long time (as Zuckerberg mentioned the company by name in his podcast with Joe Rogan) and a “war room” is just a meeting room.