I’m sorry, but unfortunately looks like we disagree in large scale.
Biased means exactly that, an expert in the topic should make the selection of the sample in order to be fair. In this case the sample is really poor. It’s really difficult to explain here in text, all the reasons why the sample is really bad. Please check an statistics course (you can read about it: here https://ocw.mit.edu/courses/mathematics/18-655-mathematical-statistics-spring-2016/ ) on ocw.mit.edu would be a lot of them.
This space is too short to explain all the reasons. I’m only can tell you that I’m doing my PhD in AI right now, and I need to do a lot of stats and the ones you show is based in a really poor sample.
About static types, again is really hard to show in this space, it would be easier in a coffee meet. However, your evidence could be found in any good book of OOP or software construction. Try Bertrand Meyer is huge but is an excellent example. So is Eiffel Language.
I will finish this disagreement with we agree that we disagree.