Actually not. This was just to measure the exact typing overhead in an equivalent experiment since this is the only language allows me to turn and off typing like this.

When working with something like Scala, I usually have the types interleaved with the coding and building just like you.

I think making this about “competency in using types” is an incorrect argument, just as the flip side “incompetency in writing tests” is an incorrect argument. What is needed are experiments to truly determine the final answer. In one of the other answers I have posted a couple of links that attempt to do just this. You should probably take some time to look through these as well.

The Unreasonable Effectiveness of Dynamic Typing for Practical Programs on Vimeo

and

A Large Scale Study of Programming Languages and Code Quality in Github: lang_github.pdf (ucdavis.edu)

Neither is conclusive, but both point to the fact that static/dynamic is only a very tiny factor in software delivery. What matters most are the people involved and how they build/deliver software.