An invalid statistical argument that appears to support the truth is still invalid. Just like a proof for a mathematical theorem can be invalid even if the statement turns out to be provably true by other means.
Nobody is "denying wealth inequality" here either, the point is just that the headline statistic may be a bad way to measure wealth inequality in crypto.
There is no doubt crypto includes wealth inequality as the network exists within our capitalist society. If 10% of the world population were to purchase crypto today, it would mirror the same wealth inequality we see in our fiat economy.
But claims that crypto networks display 100x more wealth inequality is not supported by this statistic, and articles should not be basing their arguments on this stat without understanding the implications.