If there is such a thing in the world of programming as getting a newbie to do grunt-work to help out with menial time-consuming tasks, then it is without doubt much more uncommon than in the world of art & graphics, if only for the fact that I have never even conceived of such a scenario.
You touch on an interesting subject here of "going back to your roots". In my personal habits, there is definitely a parallel to this. Whether it be in mathematics, programming, or competitive sports, I always like to keep an eye out for new people - see if I can spot someone smart, with potential. It's always very tempting to try and create a "mini me". I'm often attracted to new people who show a lot of intelligence but maybe aren't very hard-working, or very successful for that matter. But it's not like I'll high five them and congratulate them for being lazy and inconsequential like me, instead I often like to be strict with them and try to steer them away from things that went wrong for me in my own journey. 
I've had several cases like this to some extent. Usually someone 1-2 years younger than me who I'd like to see becoming my self that I hoped for. In school there was a kid who I found to be pretty smart, I'd often challenge him with math problems and logic riddles and it was very satisfying. In starcraft I had two cases of players younger than I that I tried to nourish. As for having a mentor, I don't think I've ever had one and I think I could've really used one in math. At some point I felt like Gust was sort of trying to mentor me; at some points he'd give me problems to work on, and sometimes he'd teach me random math things he found cool, before I even got interested in mathematics.
I get a very strong feel that Gust is also one of these people that wants to find a younger version of himself and grow him. He was pretty infamous for being a jerk to almost everyone, surrounding himself with a select group of people that he found worthy of intellectual exchange. Or at least that's what it looked like to me. All I really know is that he made good friends with delta.
This is also the mentality I lived with most of the time, maybe because I read Ender's Game too many times as a child. Regardless, it didn't really take off for me, I never really had more than one friend at any given time in whose abilities I had high trust.
So... in other words, with the strong emphasis on "steering them away from things that have gone wrong for myself" aside, this is also called .. the mentor & student relationship. This is something I see to be very common in mathematics. I suspect it is significantly less common in the general field of "programming" if not only because it's much less academic and more chaotic and free. and also more money-driven rather than driven by the pursuit for knowledge.
I annotated my own post because I couldn't figure out where to fit this last paragraph.
I think this should sum up my take on the "nurturing new talent" subject.