I applaud you for writing this in the way you've written it. Those who know a little bit about pharmacodynamics know that metformin, in part, activates AMPk which in turn elicits a whole host of protective effects. Moreover, metformin has been prescribed and used for decades and it's one of the drugs we have solid info on (especially when it comes to safety). This poster is a classic example of bad science...This is all, without going into the actual details and statistics of the poster, let me not call it study.
To be honest I don't think there is enough information to even say it is bad science we just don't know.
My concern is (and I have seen this happen before) that there were likely media attending that conference and they love to pick little bits of data like this (without context) and sensationalise them particularly when it is an otherwise slow news day.
I can easily imagine the headlines "Most Commonly Used Diabetes Drug Causes Dementia".
The end result may be that people stop taking it in fear of dementia with the irony that poorly controlled diabetes is one of the biggest risk factors for dementia.