The source for this article is a "study" from Chainanlysis which only tracks a handful of the possible illicit activities crypto can be used for. They are tracking scams and rugpulls, adding them together, and using that as the "illicit transactions" metric, then dividing that by a ballooned marked cap to get 0.15%.
Considering 40% of all of Bitpay's transactions are for "Gift Cards" [0] (a la money laundering), it seems to be that this blog is being as charitable as possible to CC to get their numbers.
This provides examples of phenomena that are very similar to other phenomena known to possess particular properties, and argues you don't need new evidence to deduce that there is a significant chance that the example phenomena share the same properties as the similar phenomena.
This is not congruent with the gift card example. There is no comparable product that is known to be used in the majority of cases for money laundering.
https://www.axios.com/cryptocurrency-scam-crime-popularity-r...