Chip, I'm hardly the mastermind, but I did point out a while ago, at a time when people were chanting the SRENE mantra, that true equalization was unattainable, not just in a practical sense, but in theory. When John Long started on his latest anchor book, he had Jim Ewing of Sterling Ropes test some of my predictions, which were found to be of quite practical significance in spite of their idealized formulation.
Things have headed in a funny direction since then. I think most people, once they understood that cordelette anchors distribute the load rather than equalizing it, were reasonably comfortable with that, even if what it really meant was that the pieces were loaded sequentially. But as we started thinking about (the very small number of) catastrophic anchor failures, it became reasonable to wonder whether you can up your chances of survival if you can really equalize the load.
The trouble is that you can't equalize without introducing the possibility of extension, thereby violating another of the mantra's precepts. Other tests by Ewing suggested that extension had far less effect on the ultimate peak load than people thought. GO originally pointed out that these tests may be flawed, however, and I think he is right, although I still think that small anchor extensions (relative to the length of the belayer's tie-in) are not a serious threat, or perhaps I should say that the advantages of equalization outweigh the risks of small extensions.
One of the results of that discussion was the enormous Sliding X thread on rockclimbing.com, in which various people proposed and discussed different equalizing rigging systems. Personally, I can't see myself using any of them, but others feel differently and believe that some useful solutions are presented there.
Meanwhile, I, a climber who never uses cordelettes and just ties in with the rope, became interested enough in the equalization debate to try to design something that I thought might work and also which met my standards for general utility. The initial result is the Geekqualizer (yes, GO got the name wrong), and Mal Daly of Trango was intrigued enough to sew two up for me to experiment with. I originally posted some pictures of this gadget on Supetopo; they have now been reposted by someone on rockclimbing.com as well. Neither of those threads is active at the moment; searching for "geekqualizer" or a lot of scrolling will be required to find them.
As for the pictures themselves, here are some of them:
The whole unit:
The equalizing portion is a (small) Trango Alpine Equalizer:
Set-up with unequal arms:
Close-up of clove hitch and back-up connection:
All wrapped up:
I should add that the question of relatively dynamic materials vs. newer high-strength stuff is not about improving equalization, it is about reducing the peak load to the anchors. It is fairly clear that connecting a belayer to the anchor with dyneema is not a good idea; the rope should always be the thing that transmits belayer load to the anchor. Personally, I don't think it matters that much what the anchor rigging materials are as long as the belayer is connected to them with the climbing rope.
Dawn, I don't know about fatalities, but there is one set of tests done by a canyoneering group and another set by a rescue group that indicates that falls onto dyneema (such as one might take while tethered to a rappel anchor) produce forces in excess of the UIAA maximum limit and in some cases are enough to break the dyneema sling. I'll see if I can find the references.