(Also, FWIW I personally witnessed it while trying to submit my own comment on the issue; the viewable posted comments were about 98% duplicates of that one "The unprecedented regulatory power the Obama Administration imposed on the internet" etc, and the names were somewhere in the E's.)
This wasn't real-time submission, and this should be clear from OP.
Meanwhile CQ is repeatedly insistent that the submissions be recorded individually, and not by the bulk csv method; and when the API limits turn out to be too strict they go to great effort to do it anyway: "We will need to set up multiple servers to feed the API simultaneously to meet the delivery needs of our clients" -- 114 api keys compared to one or two for most other groups.
None of which sits well with your idea that they'd intentionally sort them first. Bulk submissions might be expected to be sorted. The assumption for individually submitted comments is that they're individually submitted; it exposes the astroturfiness very clearly when all the identical comments show up in strict order in the docket. (Or at the very least, if this was just a dumb goofup or excessive tidiness, why wouldn't they have simply explained that as the obvious excuse the moment it started being reported? There'd be plenty of supporting evidence, the stage space alone would have had to have been significant)