The Gaussian depends upon a notion of independence and identical distribution: independence can be dispensed of when one thinks about power laws.

If you can soberly state the assumption that the effects of a comment on any other comment is completely negligible, or falls into a couple of other strange exceptions that leads to the universal phenomena that generates the Gaussian, then and only then can you state that any quantity associated with comments is Gaussian. I do not think that this can be soberly stated as such.