I don't know if the point is valid, but the company has made it clear in various published statements that they regard meta as not being representative of the whole user base, which includes in their opinion everyone interested in programming on the asking as well as reading side, but also answerers or curators. Meta is skewed towards the answers providers and curators and in general those who are somehow interested in this whole meta thing.
To conduct unbiased research they must either control for this skew and or find other ways to conduct research.
That doesn't necessarily mean meta is dead or not useful, just that it is only one source among many. Of course they could additionally decide to ignore it completely. They are certainly free to do that and the published post reads a bit like that.
My opinion: let them get some experience with other kinds of research, maybe the loop will make everything better or maybe next year we discuss how useful the other research is, because it may suffer from its own problems. For example 1:1 interviews may suffer from low numbers.
One example of a recent case where the meta community could substantially improve a feature: Min-Reprex: a less awkward name for MCVE
