One quick question.  Is the percentage allocation to equities a function of the current P/E10?  Recent posts seem to say yes.

Yes.

Er, two.  If so, would a retiree (doing the traditional allocation based on P/E10) 'rebalance' to a new allocation if P/E10 changes significantly enought?  That is, if current P/E10 says to go 50/50, but, a year or two down the road, your allocation is still 50/50, but current P/E10 says to be 60/40, would you rebalance to that new allocation?

Yes. I refer to this as "switching" (varying allocations according to P/E10). When used with an index fund, I refer to it as "Valuation Informed Indexing (VII or Lucky 7)." [Rob Bennett came up the phrase originally.]

I have found that training is superior to using a mechanically defined algorithm. [I had made some sophisticated calculators. I found that a human could rapidly learn to do even better.] The Scenario Surfer at my site allows you to train yourself rapidly. 

Ah, three.  If the P/E10 of the S&P500 says to be 50/50, while the P/E10 of some other fund, perhaps a value fund, said to be 60/40, would you allocate to 50/50 with the S&P500 index fund, but 60/40 to the other fund?

Yes and No. I know that I ended up using the same switching algorithms with the "Gummy Slices" of Large and Small Capitalization, Growth and Value. I do not know whether I tried to optimize them separately.

With my Current Research related to Dividends, I did change the algorithm. It was simpler. It resulted in a higher stock allocation with a higher P/E10. (I remember that P/E10=17 was the key threshold for a high stock allocation--but not the highest--as opposed to 14 or 15.) 

But remember, history does not repeat exactly. I have run sensitivity studies. They show that missing the exact optimum is OK. The optimal thresholds and allocations are quite broad.

Have fun.

John Walter Russell