There's a slight argument I would say, regarding the methodology to generate those US ASL figures, if one takes a moment to think of it. With most languages listed, people self-identify with a corresponding language and base this answer to the census/polling origination. The individual doesn't need to be verified by a source, they can (generally speaking) personally assest if they are capable of declaring it as their official language. This would incur for many of those examples, anyone speaking of Spanish, Chinese, German, etc. languages.
However, as with ASL isn't there generally an unspoken sub-requirement in that you are accepted into the community before an individual can declare oneself officially a part of it? These participants of the Deaf community tend to judge the individuals.
Sort of like the college example given, taking a couple of college classes in Spanish doesn't automatically constitute an official "declaration" of sorts that you are now a native speaker of the language, you personally have to assest this yourself. Except this time for ASL, it is done for you.
The issue is for ASL, I'm trying to convey, that it seems to be the other way around - they [the community] generally are to accept the individual before they can officially declare themselves apart of the pool. In this effect, that would actually lower some amounts of people in that group - impacting mostly hearing that declare their language(s) with ASL. On the flip side, in effect, it could also increase the likelihood of people who associate with it because they are deaf, although not fully fluent in the language thereby generating extras than necessary.