[thstone]

The best way to know what you have is to perform a bone stock dyno test and then re-dyno after every mod using the same dyno to minimize variability. Temp, pressure, etc, will still make even those comparisons hard enough.

Unfortunately, trying to use someone else's loss estimate or dyno values to calculate loss is just a best guess. Why do I say this?

I have seen various Spec Boxster stock 2.5L engines tested on the same dyno on the same day vary from 172hp to 188hp. That variability is larger than the hoped for horsepower gains from most mods. That is why Topless said that this is a fools errand - the error is larger than the value being measured.

Moving forward, get a good baseline and then compare that to any future mods. Trying to build the horsepower history backwards is generally an unreliable approach.