I know a an accomplished CS professor, ACM fellow, cited in Knuth's TAOCP (as well as being an easter egg!), who still hunt-and-pecks. In fact, hunt-an-pecks incredibly slowly.

Seeing him type really reinforced this idea.

I've always told my Jr Engineers to "think twice, code once".

If I gave them a task and they immediately started typing it out, I would tell them to stop typing and ask them to explain to me what they were doing; they'd often just spit out what they thought the code should do, and I'd often point out edge cases they missed and would have missed had they just spit out code and a PR, wasting everyone's time. I would also insulate them from upper management to give them time to actually think (e.g. I wouldn't be coding so they could think then code).

To your point and to the GP's point, and one point I keep raising with LLM's: "typing is not where my time sinks are"

That's very true, which is why I find it insulting that so many AI proponents use the word "typing" to refer to writing code. It carries an implication that if you enjoy writing code by hand, you enjoy a mindless activity.