Because regulatory comments aren't a vote or popularity contest, or even subject to much public attention, but are instead, in the normal case, subject to substantive review and response, ballot box stuffing in a way which makes them appear to come from less expert general public commenters doesn't really help the corporation. It only really helps on an issue where the regulatory process has public attention where the substantive outcome is a foregone conclusion but wants additional public political cover. This is quite rare for federal rulemaking.
(A milder—in volume terms—version where you try to make substantive comments seem like they are coming from involved parties with a different interest than yours—say, on a healthcare reg, a payer presenting fraudulent provider comments, might be useful, except in that case there is a lot more risk that someone would follow up on the particular comment and turn up something fishy.)
I think the recent tax cut is an example of a process with a predetermined outcome where politicians were seeking political cover to make a vote the majority of Americans disagreed with. It’s certainly possible that bots on twitter, Facebook, Instagram and reddit could be employed to amplify a particular message and provide that cover.
(A milder—in volume terms—version where you try to make substantive comments seem like they are coming from involved parties with a different interest than yours—say, on a healthcare reg, a payer presenting fraudulent provider comments, might be useful, except in that case there is a lot more risk that someone would follow up on the particular comment and turn up something fishy.)