It's rare when I agree with helonorth. The temperature in hell must be awfully cold right now, because I think helnorth has the best handle on this VRS thing and how to extricate oneself from it. Look, fully-developed VRS is not fun. As Crab notes, you have almost lost control of the helicopter (except for the tail rotor, if you have one). 1) If you're up high (e.g. external load work), then do whatever you need to do to get air flowing laterally through the disk. 2) If you're down low and don't recognize it in time, you're probably going to crash. And if you're too late in employing that Vuichard technique, you're probably going to crash going sideways.
VRS events that result in a crash/hard-landing usually don't happen way up high (duh). They happen down low at the bottom of a screwed-up approach. If you're even halfway awake, you can catch the VRS while it's still in the I-VRS stage and bail out if you can. But if you're already close-in and have maximum (or nearly so) power applied...well. The BO105 was horrible in this regard. With respect to that "R-44 rooftop scenario"... at some offshore oil installations, due to the placement of the helideck, you just cannot get into the wind on final. So you'd sit there, downwind (more or less) with the machine vibrating like crazy (as all Bolkows do as their rotor passes back through ETL) and your left arm is finely tuned to any "disconnect" between collective movement and rate of descent.
Just as we don't do full stalls in swept-wing jets anymore, real VRS is not something that should be practiced. It is something that should be avoided.