Clive, thanks for a great spreadsheet and example.
My back-testing pretty much agrees with Clive's, and the delayed method performed significantly better than BTB, both in terms of risk and reward.
My BTB results agree 100% with Clive's.
My Ocrfot's delayed-AIM results are a bit different than Clive's. The reason is that I followed Ocroft's steps exactly how described: buy only when AIM BTB is recommending no buys at all. So in Aug 2008 for instance (SSO price of 58.85), even though BTB AIM stopped buying, there would've still been a small buy order (that obviously did not meet the min amount) and the delayed method did not deploy any money. Eventually, all accumulated buys (value of 60,810$) were executed in March 2009 as the tooltip shows.
Here are the results. Note that at the peak, there was a difference of ~100,000$. This a major improvement.
News 
Market Data 
Discover 
