Take a look in ANAM at the Tahquitz double fatality where a cordelette caused sequential failure of a three point anchor because it did not equalize the three pieces. Sliding X's or an equalette would not have failed, as simulated replications of the accident using the same gear that failed in the same placements and test weights showed
It seems this was an error in setting up (equalizing) the anchor, not a problem with cordelette. The reason I do not use the Sliding X is because if one anchor point fails, the resulting extension will likely cause the others to fail...as in this case of a poorly equalized anchor. And, of course, the reason we use multiple anchor points is b/c we do not assume any single point will hold a fall.
Am I mistaken? Should I be reconsidering using the sliding x over an equalized and tied cordellette? In what situations?
Just read the Supertopo post for details if you are interested, but in essence, even if a cordelette LOOKS equalized, it transmits the majority of the force to the shortest arm, resulting in asymmetric loading of the pieces. Also, it is non-directional - if you move even a little bit, it unloads one arm and maxi-loads another.
In contrast, it turns out that when tested in actual failures, even if one arm of the Sliding X unclips, the friction of the X sliding, together with stretch from the nylon, results in no shockload on the anchor.