My experience with trailing stop losses is different.  I trade using Fidelity's system.  Typically I use their Active Trader Pro.  With Fidelity, I can establish a trailing stop loss at a particular percentage off the purchase price.  Fidelity adjusts the trailing stop maintaining that percentage within the day if the stock moves higher.  It's a real time adjustment as best as I can tell and effective within the trading day.  I'm not a day trader, but I appreciate that I get the benefit of intraday moves to my advantage.

My backtesting system is another matter and it is based on historical end of day data.  The system allows me to try various exit strategies including trailing stop losses based on percentages (similar to Fidelity).  My backtesting works well for trailing stop loss adjustments when the stock is held for longer periods of time - say greater than 30 days.  But the problem I was trying to define above deals with 'simulating' intraday moves that affect lower trailing stop loss percentages when all you have available is end of day data.  My system currently penalizes results by only taking into account the open or the low being below the last trailing stop loss.  It doesn't take into account the possibility that the stock may have hit a new intraday high and then dropped to the low.  So I tried to average the intraday effect and account for that in the results.

I realize that there may be no solution to this; that the only solution is to be backtesting with intraday data...but that's not possible for extensive testing over long periods of time.

I hope this helps explain further.