I do think this is more an issue with science communication broadly than string theory specifically - every field has its own examples, and medicine is notorious for it - but she is right that scientific researchers (the subject matter experts) have a responsibility to accurately communicate their work when speaking to the public.
Its one thing for an enthusiast to inadvertently oversell a concept to the public as fact because they are excited and only understand at only a basic level. It’s another entirely for someone who’s been researching that concept for 30-40 years, with the express intent of proving or disproving its validity, to oversell it as fact when they’re whole job is to be intimately familiar with its shortcomings. They, of all people, should know better - and that means they have a responsibility to do better.
Science does get messy, by design, but it is the duty of those who communicate their science to be honest about that messiness, not mask it by unfounded statements to sell their ideas to people that don’t have the research expertise to spot the falsehoods.
Largely, yeah, we just don’t know. In research terms, it’s still quite early, so anything definitive is likely years away. However, research is starting to indicate it may be neurological - a product of damage to the brain and nervous system, as this article discusses.