I may be considered "old" and have worked on the 47 a little. The "violent reaction" is the fact that apart from the starter trying to engage (messy) you will have energised the "shower of sparks" and the retard points in normally the LH magneto. For the retarded i.e later ignition to be effective the normal points are shorted to ground. Engine will go a little funny all right. I guess they don't call them "retard" points for nothing!
Robinson setup is pretty much similar.
Have a read
http://www.aeroelectric.com/articles...erOfSparks.pdf
The key switch controlled diagram is the one to look at.
Probably the single biggest reason for starter failure is the poor condition of the starting circuits in the mags etc. BTDT. If it doesn't start at any time within a couple of compressions it CAN be fixed!
The practice of cranking with the MAG switch in the OFF switch and then switching to BOTH (while still cranking) would probably instigate physical violence from most engineers witnessing the event. If you really want to crank to check for "flat" cylinders, which is an OK practice, do it with the MAG's OFF, STOP CRANKING, SWITCH MAGS ON, then start the engine. Yes I am YELLING.