I'm not going to waste my time either, reading the entire article as it's been my experience reading similar comparisons that percentages are often used without a solid platform explained...at least, not explained well enough for my feeble brain to understand.
I have a problem with percentages when what I really would rather read about is actual time elements. If something like this were to tell the users that Firefox slows down by...let's say, 3 seconds when using whichever add on, then I'd be able to run my own comparison using that same add on.
It's my gut feeling that these percentages are most likely relative to a variety of wavering factors that also most likely probably most users wouldn't encounter.
Anyway...if my Firefox is slowed down by three thousandths of a second because of an add on that I find useful...or even if it's slowed down by three full seconds or so, I'm not going to run to the medicine cabinet to grab the tylenol.