Joined: Nov 2016
The "100 lbs = 1/10 of a second" is only a very vague guideline. On a heavy car (let's just say a '70 Cadillac for example), you can scarcely tell if you've added a passenger, much less taken 100 lbs out. It will likely affect you less than 1/10 of a second.
If your car's around 3000 lbs, it's going to be pretty close.
If your car is very light (think sand rail) then removing 100 lbs could well result in far more than .1 second.
It's also important to realize that where the weight's at, and transfer, can really affect traction. If you pull all the weight out of the back end, wheelspin may take away any ET gains. There are lots of fun Youtube videos of people that take this to extremes. There's a great old Roadkill episode where they strip down an early 80s corvette and make it genuinely fast. There's also a much more recent episode done by the Hoonigans where they strip a police car to find out if it'll be faster. (Spoiler: It mostly goes slower, until the very last step of weight removal, due to traction issues)
Also, it's interesting to note that rotational weight affects this more. So, for example, you put on a set of magnesium Halibrands and save 10 lbs per corner of your car. That's only 40 lbs, right? But because it's moving, the inertial effect is much greater. You might gain over 1/10 of a second with such a change, despite the fact that it's not 100 lbs. This is why most serious drag cars have the smallest, lightest front wheels and tires they can get away with. It definitely has an effect.
Lightened driveshafts have only a small effect, because of the way torque works. (remember the grade school lessons about 1 lb weights on the end of a 1 foot arm, vs. 2 foot arm?) However, at high speeds with overdrive, they can definitely reduce driveline harmonics and vibration.
Engineering. It's complicated. Who'd have guessed?
I smile a lot. It makes people wonder what I'm up to...