Yes it is way better! If flat Earth is so outlandish (which it is), it should be easily dismantled in debate and journalistic investigation. Not wanting to allow it full stop for fear of internet trolls grouping up is no solution at all.
Sure: journalistic investigation should quickly dismantle it, which means they shouldn't report it to the public. It would be like reporting someone's words that you have found are lies, just to be fair and balanced.
On the other hands, debates with a general audience (i.e. non-experts) are a terrible way to judge scientific arguments. Non-experts are simply unable to judge for themselves the strength of most scientific arguments, because they simply don't know the fields well enough. The vast majority of the arguments will fly by us and we'll be left as confused or more; and we'll pick winners based on charisma rather than any deeper thought.
Yes but people must do that every day anyway. It's like saying no expert opinion or statement is valid unless you check credentials first and they meet some arbitrary bar of merit.
And even then, we've witnessed a long history of 'credentialed' people being corporate puppets for certain agendas, peddlers of their own self-importance and fraud or just plain wrong.
We should pick experts, but we should pick from a range, not just the experts we like or decide are on the right side. And sometimes also pay attention to the non-experts, they can ask very interesting questions and provide good thought experiments, even if naive.
We should pick experts that other experts believe in. If we happen to be experts in a field, or even adjacent fields, then by all means, let's pick experts based on our own understanding of their arguments. Knowledge does transfer to some extent, so if you're a mathematician or physicist or statistician, you could probably judge some climate or medical papers based on their statistical methods - especially when you find flaws (even if the math is perfect, if you don't know why they chose a particular model and how well it might actually fit the thing being modelled, you may not be sure they are not choosing a known bad model).
But if we're not, the only thing we can meaningfully do is to look at what other experts are saying is the right opinion, and judge based on popularity in their field.
