Agencies must consider the comments, but as 'JumpCrisscross points out, they don’t have to accept the commenters’ views. The outcome is determined not by submitting the most or best comments, but by electing Senators and Representatives to enact statutes defining agencies’ authority and responsibilities, and by electing a President to appoint their officials.
If you’re new to administrative procedure, you might find it interesting to read though the entire dockets in some less-notable proceedings, to get an idea of how it usually works: an agency issues a notice of proposed rulemaking; interested parties make comments (and sometimes reply comments) with their views, data, analyses, and legal arguments; and then the agency issues a final rule, usually explaining how they considered and accepted or rejected the commenters’ positions.
It’s rare for comments to cause a proposed rule to be abandoned entirely, but they often influence the myriad details—definitions, exceptions, compliance deadlines, technical specifications—that make administrative regulations a practical tool of representative government.
Limiting input to that which addresses procedural questions or point-of-order effectively narrows conversation so that other perspectives outside of that window is made irrelevant.
Effectively pushing the public out of policymaking. (Good riddance to all that white noise!)
Political appointees are entirely within their power to ignore such comments.
So an actual analysis (which would necessarily include predictions that are not certain) of the impact of a decision is irrelevant?