A survey essentially asking people "do you have long covid?" isn't extraordinary evidence. Could be a simple case of psychogenesis[0] like has been common during any mass illness event historically. The demographic info in your link seems to indicate that is a possibility. With 25 percent of the population apparently having this disease it's not a hard one to study experimentally.

[0]https://en.m.wikipedia.org/wiki/Mass_psychogenic_illness

Except that we have all kinds of other indicators of this phenomenon as well.