As far as I know, the controversy involves a pair of mathematicians using LLMs to solve the problem and another group who used LLMs to solve the problem. No one seems to be disputing the solution itself.
The controversy was that OpenAI may have stolen the work of mathematicians, which IMO is not hard to believe given they've stolen quiet alot for training data.
Right, but those mathematicians whose work might have been stolen were...using LLMs to attack the problem.