The claim (not directly made by the article, but generally made to support the difference) is that vinyl’s physical characteristics limit how “loud” a recording can be. Specifically, that a vinyl pressing of a “brick walled” recording becomes unplayable—or at least unreliably playable—because its physical tracks are insufficient to keep a stylus in place for playback.
I don’t know how true that claim is, but the analysis seems cromulent, and analysis of comparable media seems to support it well enough.
It isn’t a claim that pressing vinyl attracts or requires better production etc, but that the medium has inherent physical constraints that benefit, at least as a side effect, from greater dynamic range.