George Stebbins' full accident report of his unprecedented harness failure at Kagel, an excellent example of dealing with the unexpected--and surviving:
Master rated pilot with about 1700 hours launched borrowed Moyes Litespeed S5 in no wind on sled-ride day. Harness was Rotor KickAss, inspected within last 6 months. Launch was fine. After fully clearing hill and a few mild maneuvers, pilot put glider into full dive with 100% VG. Smooth, no turns, no problems. Pilot began to slow down very gradually. Altitude was between 1000 and 1500 feet AGL.
After two inches or so of slow-down, pilot heard loud "pop/snap" sound. Pilot's first (and incorrect) thought was that the cross-bar had failed. Pilot fell violently, and the base-tube flew out of his hands, slamming forward into his throat. Pilot is unsure if his hands moved forward with the basetube part of the way before he lost his grip. The pilot re-grabbed the basetube. At this point the sudden pitch-up of the glider had started a loop. The pilot had grabbed the base-tube off-center, so the incipient loop turned into a wing-over. The pilot managed to get the glider under control, but was hanging vertically (in landing position) and his forehead was approximately level with the base-tube.
Pilot looked at the parachute handle, and at the LZ and glider. He then looked at the harness straps to see what had happened. It appeared that the harness mains had failed or come detached from the harness, and pilot was hanging from the integral backups in the harness. Pilot could not tell, but wasn't sure of the integrity of the harness. This made deciding whether to deploy or not difficult. Final decision was that it might be dangerous to land, but it was more likely to be dangerous to deploy. Glider was intact. Harness did not seem in immanent danger of further failure, and pilot was familiar with the LZ. The decision was to attempt to land.
Pilot let go with one hand to unzip the harness (thinking that he'd rather do this high.) This caused immediate and severe oscillations, but pilot did get harness unzipped enough that his legs could finish the job. Pilot then got glider to stop the oscillations. (Note: This was a clue that the glider was flying very fast, but pilot didn't catch on to this yet. He assumed the over-control was due solely to hanging so low.)
At this point the glider was nearing the LZ, and at about 800-1000' AGL. Pilot set up an aircraft approach that was standard for him at that LZ, being careful to use shallow turns. But he seemed to be sinking far faster than expected. Pilot realized that hanging that low, the geometry made his farthest reach too short to keep the glider anywhere near normal approach speed, but didn't realized just how fast he was going. As he got 1/3 through his base-leg, he realized he wouldn't make the LZ in this pattern. So, he started to turn directly towards the LZ to attempt to land in the opposite direction than normal. A "downwind" landing seemed safe since there was no wind to speak of. This turn caused pilot to drop further, and he was now not going to make the LZ even going downwind. It should be noted that as the pilot got lower he realized that in this position, his vision was restricted to "just below" horizon level and upwards. In other words, he could not see well below him at all.
Now pilot was in trouble, since the bail-out LZ was full of boulders....
Continued at http://ozreport.com/pub/rotorfailure.shtml
[For Wills Wing follow-up investigation and subsequent Rotor harness mods, see http://www.willswing.com/news.asp?newsSeek=228
An excerpt of general interest:
" ...pilots should be aware that a sudden transition from VG tight configuration to VG loose, while flying at high speed, such as could happen if the VG rope slipped from the cleat, will usually cause a severe pitch up that the pilot will not be able to adequately control, and that this can result in a very dangerous flight condition. "