The sample size is statistically correct (see for example
https://www.surveymonkey.com/mp/sample-size-calculator/), though it always seems a marvel how such a low amount of people can represent the whole population of a country in a survey. The most important factor, in addition to the sample size itself, is the adequate selection of the candidates to avoid biased answers (i.e. gender, age, geographical distribution, education, randomness, and so forth). Although the footnote on
Pantera’s site claims that the sample is representative in those terms, there is no evidence (i.e. description of sample distribution by these factors, method of selection, method of survey).
Subjectively, the 22% figure seems like a stretch to me, and I’d be surprise if there wasn’t any bias in the randomness of the sample and/or survey method (i.e. if it were to be online, that would cause a bias per se).
One thing to stress is that the survey really indicates 22% of Belgian employees are ready to accept
partial payment in crypto. I stress the word partial, which is a relatively open concept. I’m pretty sure that, were there to be a subsequent question querying percentage ranges, the results would be more prone to be in the range of the lowest percentages.