I have never seen my PhD mentor or postdoc mentor perform experiments by himself. Most established PIs do not do experiments by themselves any more, although I imagine they used to be great at it. They are too busy. I am not that too busy, and not established either, so of course I still do experiments by myself.
However, I am getting busier with things other than experiments than when I started the lab for sure. I have more paper works, grant dues, committee duties, and so on. I cannot still imagine leaving the bench now, but guess eventually I will. So, the question is when?
When I was a postdoc, I heard multiple PIs advised that starting PIs should stay at the bench and do experiments as long as possible to set the tone in the lab. I agree with that idea, so I try to do so. I took the UCSF-Gladstone Scientific Leadership and Management Skills Course. (I highly recommend anybody to take this course by the way, if you have a chance.) One thing I really remember is that the lecturer said "You would like to think you are special, but you are not. Most of things you are doing can be done by someone else in the lab." I think he meant for some lab duties or something, but I feel it is sort of true for experiments as well.
I used to be the best at performing ddPCR experiments in the lab, but now people in the lab can do that very well. I am probably still the best at isolating genome edited iPS cells in the lab, but that will change soon, as I am training them now. They are doing a great job helping me out. Actually, it is great that everybody can do what I do. That means that our work is reproducible. This is the way science should be. At the same time though, many of my bench works have been taken by someone else in the lab, and I do definitely a lot less at the bench than when I started the lab all by myself.
I guess this is a part of growth of a PI, but I am not leaving the bench (yet). Let's see how it goes from here.