Maybe I'm just slow or dense, or both... but there doesn't seem to be any real "exact science" to a 100% crop. You took a full res image and used the marquee tool to make a crop box, which if someone else had done it, could have been a different size. You then cropped the image and saved the selection as a new file.
I guess what I'm beginning to understand is that one guy's 100% crop is going to differ from another and there really is no exact science to it when you select a section with the marquee tool.
If my theory is correct, could one create a new blank canvas, maybe 100x100 or 200x200 pixels and paste the full res image onto that canvas? Once the full res photo has been pasted, you could move it around until you find the section of the photo you'd like to keep as your 100% crop.
This way, all of your crops will at least be consistent. Should the amount of pixels differ from a shot taken by a 30D vs a 5D, considering the megapixel difference?
Honestly, the term 100% crop doesn't make any sense to me. Defining something with a percentage generally allows there to be an equation that can be used to always obtain the same result. A formula. There doesn't seem to be any formula to go along with this "100% crop". To crop something, means to take away, right? So, if I have a number, "50" for example and I take away "100%", I'm now left with nothing. Wouldn't it be better to do a 5 or 10% crop, that way you have a consistent amount of pixels?
Sorry if this has been answered/addressed, but after pouring over the threads, I can't seem to find a straight answer.