Assuming a diver scooters at 150 fpm and swims at 50 fpm, it's obvious they'll need at least three times the time to exit via fin power. What is slightly less obvious is that the swimming SAC is going to be quite a bit higher than the scootering SAC, so you'll need a lot more than 3 times the penetration gas to safely exit. Assuming you use *only* 66 cu ft of penetration gas from 2 stages, and have a scooter SAC of .4 compared to a swimming SAC of .6, you're looking at around 300 cu ft of gas needed to exit if the scooter fails at max penetration in a no flow cave.
If you consider the gas in a pair of 130s as well as 80 cu ft of gas in the two half empty stages, you've only got 340 cu ft in reserve and you'll need 300 cu ft to swim out even if no other failures or delays occur. That's also assuming you don't rack up a significant deco obligation on the 3 times longer exit.
Consequently, it might make sense to drop the first stage (or not depending on the depth, flow, scooter, cave, etc) but the second one will stay on you as you'll turn the dive when you hit 1700 psi on it, as you probably don't want to touch the back gas on a two stage scooter dive.
Add a faster scooter in the 200-250 fpm range and the gas requirements for exit go up yet again as you are now swimming at only 1/4th to 1/5th the scooter speed and the reserve needed to exit becomes 400 and 500 cu ft respectively.
Where it gets even more complicated is when you scooter into the cave and then swim farther past the point you drop the scooters. You then have to figure the impact on your reserve gas as well as the rock bottom needed to swim out in the event of a scooter failure plus the potential deco obligation.