Some of us used to play computer games (on the 286 machines), while our programs ran on the mainframe. Once the mainframe program completed running, in about 20 minutes, used to go and see the results, and debug.
Depends on how 'far-sighted' your employer is - there's been several occasions where knowledge I've gained from (strictly speaking, 'private' study) have saved the bacon at work.
Try telling my boss, and I'd probably be censured for 'wasting' work time...
It depends a lot upon what you do as "your own projects".
In my chemist days it was the same. I heard about predator/pray simulations and happened to have read a paper about ligand substitution on Molybdenum Hexacarobnyl.
So I went and amused myself with my first Monte-Carlo model.
My employer has the patent that resulted from it (and eventual experimental proof) "Induced Surface Ensembles on Transition Metal Surfaces".
I guess it depend upon how one plays.
On the other hand - it seems all the "clerks" (and above) have phone on there desk that, when not streaming videos is otherwise sucking up about half their time. In this case, it's not even productive for some third party. That's a problem that's real and pressing.
if you are getting paid money to work on your private projects,
That I was not doing.
Building a website for a (really) non-profit and maintaining it, for example.
Others are just plain raw curiosity - like ODBC on Excel and Text files. Sure enough (actually, amazingly) the need to extract and manipulate Excel on its way to SQL occurred in less than two week. An import just didn't work for it.
I do agree that "freelancing" on company time, or billing A whilst working on B's project are unethical practices. Stealing is an appropriate description.