Twitter in particular is one of the cases where places like HN disappointingly fall for the fallacy of assuming that their observations are an unbiased sample (and thus that "most/tons of people use Twitter"). Twitter's numbers have always been shockingly low (compared to their influence); I remember finding out a couple years ago that active members of the narrowest possible metric of actual social usage of Google+ roughly matched or outpaced Twitter (this was before many of the integrations like Youtube which may have drawn more people to the G+ stream). That was the case for quite a long amount of time.
It's probably great news for Twitter that their user numbers so badly lag their influence, since that means the potential for growth is very, very high (though it does raise the question why this gap exists to such an unusual degree).