This is pretty much just dollar cost averaging, right?
This process is a 'rate capture' process, which is that a fixed amount of principle captures a moderately proportional fraction of the change in price. With a capture limit (in this case 50%)
I built a simple spreadsheet for you [1] which can illustrate this. In the spreadsheet there are two scenarios, one the person "pulls the trigger" every time their investments have grown by the target amount, pull off the growth and take it out of the game, and then go back in with their basic investment. In the second scenario they just let it grow and grow and grow. After 5 rounds with a growth target of 50% they have $35K if they regularly pull the trigger, and $75K if they let it ride (more money is in play so the same percentage increase in the market gets you more money back). When the bubble bursts, if you lose all of the money in play, you end up with $25K and $0.
You can play with the spreadsheet and make some other choices to see how they work out. The smaller your trigger point the closer you get to having identical returns because you're not getting any amplifying effect of the money allowed to grow, the larger the growth target the longer in time (generally) you have to wait before you can pull the trigger and like musical chairs find yourself losing out because you waited to long.
An interesting diversion is to run different investment strategies against the historical record from the bubble to see which left you with more money. If we are in another bubble that could be useful information.
Or not. See other posts on the randomness of it all.
[1] https://docs.google.com/spreadsheet/ccc?key=0Atwe7dq6iPQHdGx...