Do you go fishing, if so, you've opened a can of worms and should hang out for some free bait.

First of all, as I mentioned earlier, this is simply a rough approximation of net effect of weight change. So bike or body, it's all the same thing.

OTOH there's plenty of opinions, and some science, on the differences between weight changes on the body, bike or wheels. Lot's of debate on the subject, comparable to a chain lube debate, and few meaningful answers. There is a general consensus, that weight reduction at the rim counts double (supported by good science) for purposes of acceleration, but not for climbing or constant speed riding. There's also some (not as much) consensus that weight on the bike counts a bit more than on the body, reflecting the fact that riders move the bike around under them (I buy into this, but consider it to be subtle).

Also, think about what happens when climbing steep grades at low speed and cadence. The bike undergoes decent size speed changes with each pedal stroke, so the effects of weight are slightly greater than at constant speed. Then there's "feel". Lighter bikes feel lighter and snappier, that may produce a placebo effect that helps riders ride as if it actually made a difference. IMO- for whatever reason, losing 5#s off a bike feels like a bigger difference than losing 5#s off your body.

Throw all of this into the hopper, and draw your own conclusions, and maybe tweak the overall graph a bit, but it's just another set of small variables so don't assign it more value than it deserves.

The rule of thumb here, and in similar discussions, is that the science can get you to the right city, and even (maybe) to the right ball park, but if you think it can get you any closer than that you're kidding yourself.