You can still say type systems are obviously a good thing (I do too!). But you, sadly, can't say this is supported by scientific studies. It's an opinion based on your own personal experience.
The good news is that this opinion doesn't contradict evidence either! The research just isn't there yet. Some of research I've seen so far is pretty bad.
The fact that you actually recognized this response in yourself means that you're more than capable of adjusting that response to be less biased. It's a very good first step.
One option is to shrink your world to the things you have direct experience with. So yes, type systems are obviously a good thing - for you. Then the obvious question is why would somebody believe otherwise - and you can learn something by asking it. Worry about figuring out what the world is actually like once you've got a sample size of more than two.
When enough people are aware of the meme, and the meme is ingrained in culture enough to have a statistically significant effect on perceived ability. Or when people decide it is meaningless, out of desire to control their expectations of their performance.
The first three I was all like "yes! So glad someone articulate is finally saying this".
Then for type systems I was unable to overcome my current biases, thinking "what a load of tosh! Type systems are obviously a good thing".
Makes me wonder what it would actually take to break the Dunning-Kruger meme