My logic goes like this:
If you put significant effort and time into creating an equalized anchor, then it is because you feel that there is a 'significant' chance of needing that equalization. I'll assume 3 pieces, and perfect equalization.
If I don't clip, a FF2 (assumed 9kN) goes onto the anchor, or 3kN per piece.
If I clip only one piece and the leader falls, then that piece sees 15kN, and when it fails, the remaining two see 4.5kN each.
If I clip the powerpoint, each piece sees 5kN.
Conclusions:ÂÂ
One: For high FF falls, the only reason to clip the leader in is to make the catch easier on the belayer, because it always increases anchor loadings
Two: If you clip, you should clip either the powerpoint, or a separate piece not part of the anchor, because you must be confident that the remaining pieces will hold - therefore that piece was not needed in the anchor. Not including it means you don't have to worry about how your anchor will equalize after failure.
All of this assumes a lot of worst case things, but that's safety engineering. Also assumes some best case things - equal quality pieces, perfect equalization, etc. But I think all high impact cases point to not clipping an anchor piece. But in practice, of course, it depends
For example, say the anchor pieces are pretty high - 10' off the ledge - you could be looking at a 20' factor two vs a 4' factor 0.4 if the leader slips up there. If he expects to get another piece in before the FF will get high, of course he should clip a piece, as much to limit his fall as anything else.
And on the third hand, if you got bomber stuff and aren't relying on equalization, then why not?
But in the grim situation - questionable pieces, good runout coming up - take it on the waist!