Actually, the compressed gas in a CO2 bottle takes a fraction of a second to activate a weapon (for all intents and purposes, immediately after the match starts, the weapon can be used). The pressure drop of the escaping CO2 in the process will cause it to cool, but the energy is immediately available regardless of this cooling. To prove my point, consider that you vented all of your pressure in one strike/flip/etc. This could be done immediately.

A gasoline engine cannot burn up all of the chemical energy stored in its fuel tanks all at once and transfer that energy into the opponent. A battery cannot drain all of its power through a motor in a single, quick burst either.

The gasoline and battery powered systems are limited by their horsepower or voltage/current limitations, respectively.

The appropriate analogy between a CO2 system and a gas or electric motor system would be entering the match with the chemical (gasoline or battery) energy pre-transformed into immediately usable energy such as kinetic energy (spinning disk) or potential energy (compress a huge spring such that it stores kJ-order energy).

As for the amount of energy that can be stored in a flywheel, if you get a dense enough material positioned at the maximum radius your design wants to armor (tungsten is a very dense possibility), and you connect this with a lightweight set of spokes or slotted disk, you can store a tremendous quantity of kinetic energy at a few thousand rpm. You dont need 100000 rpm.

Did some scavanging to collect the info, and if you had a tungsten ring that was 3cm thick, had an inner radius of 37.5cm and an outer radius of 40cm, it would have a mass of 35.1kg. If you then spun this ring up to 1000rpm, youd have 29kJ of energy. If you dumped all of this into a 100kg bot straight up, itd go 29.5 meters straight up!! And what goes up...

This is a bit unrealistic, because it would take quite a long time to spin this disk up to 1000rpm, but you get the point; this would fit in a bot.

I have an idea of how to get the energy out of the disk into whatever needs it, but I have to do a bit of research to determine if my approach would work...and then there is the minor detail of actually building it ;-)

Mack