To preface, I would like to hear thoughts on the below first, before anyone takes action on it. I'm aware that there is a time and effort crunch with GSoC involved.
I think the overall issue is using this Discourse feature over a shareable offsite survey platform, like in BibTeX Plugin: Survey.
Joplin Discord, subreddit, and other social media users should be allowed and encouraged to take the survey and provide their feedback.
Even then, a survey like this can be completed in good faith by those unaware of Joplin. Partly to Laurent's point, those could include IA, AI, and notetaking people. They may understand what makes for the best summaries better than we do.
I think using a survey platform as above, combined with shorter articles and summaries, would yield many more participants and thus better results.