I've never seen a satellite torando but have often seen more than one tornado on the ground at the same time. I'd say they are pretty rare(actual satellite tors not the more than one on the ground at once events).
I have seen satellite tornadoes, all of which were on May 29, 2004 on the Beloit-Belleville storm. They formed on the periphery of the violently rotating mesocyclone and appeared to be rather weak. Sorry I don't know more about how they form, but I would say that before we can distinguish between how different "types" of tornadoes form we will have to gain a better understanding of how tornadoes form in general. Perhaps there is someone else in the group that is a little more versed on current theory, I have been out of school for 10 yrs now and I know that the science has come a long way since then.
I know there was a satellite tornado on the May 12, 2004 Harper county, Ks storm. I'm pretty sure it was not associated with the Attica Tornado though. Maybe the next tornado after that? I did not witness that storm so maybe someone who did could specify on that.
I got a real good look at the satellite tornado on the May 29, 2004 storm because the damn thing touched down about 50 yards in front of my car and scared the hell out of me. It was one or two tornadoes after the Attica tornado. I don't know if you could really call that a satellite tornado. It didn't last very long so you couldn't really tell if it was orbiting the bigger tornado.
I think true satellite tornadoes are pretty rare. The only one I have ever seen was on the Hallam tornado and on the 5/29/04 Harper county storm if you count that as a satellite. I am pretty sure the May 3 OKC tornado had a good satellite tornado for a while. I remember hearing Wurman say something about how common smaller circulations are around the periphery of the mesocyclone, so I would imagine that there are quite a few spinups that result from these smaller circulations, but if they don't persist for very long they probably never get credit for being a satellite tornado. I guess it just depends on what the definition of a satellite tornado is (which I don't know). I would think you could probably find some stuff on the internet from Wurman's field research that would kind of cover this topic.
Satellite tornadoes orbit and form near the main tornado (usually a strong tornado/mesocyclone) and from the same mesocyclone. It's possible Saul that what you are refering to was in fact a multiple vortex tornado, where several sub vorticies are rotating in and around the main larger tornado.
There is also the process of cyclic supercells, there were several cases of this on March 12th. Where one updraft/mesocyclone would weaken while another one formed and strengthend. The end result was the previous tornado was still on the ground while a new one was forming and in one instance from the Mar 12th Mega-cell. It had one meso with a main tornado/satellite and a new meso with another tornado on the ground all at the same time!
I don't recall ever seeing a "true" satelite tornado, but I recall numerous instances where there was multiple tornadic circulations on the ground (particularly on 5-14-2004, 6-9-2005 and 6-12-2005). Here are a few caps from the cyclic tornadic supercell on 6-12-2005 in northwest TX (the "Jayton" tornadoes):
These circulations were obviously produced by the same low-level mesocyclone.