Thank you Crissa.
Volume of a cylinder is Pi x R
2 x H. So for 4680 cell it is Pi×23
2×80 = 42320 x Pi = 132952 mm3, and for 18650 cell it is Pi×8
2×65 = 4160 x Pi = 13069 mm3. I used this website for the calculations
https://www.calculator.net/volume-calculator.html?The ratio can be done in many ways but without grabbing a calculator it is easiest if you half 132952, then you get 66476, if you half that again (1/4) you get 33238, if you half that again (1/8) you get 16619. So 1/8 of the 4680 cell is a bit bigger volume than 18650 cell.
But I did go to the website again and divided 13069 into 132952. It said said 13069 would go 10.17 times into 132952.
So it was slightly worse than I thought after the quick calculations I did in my head. Tesla gets 5-6 times the energy out of 10 times the volume. Now, this math must be wrong somewhere. Why would they start making much bigger cells if they only get half the power out of the volume? Partial answer to that could be simply fewer cells in a pack. That makes the BMS simpler and cheaper. Maybe they'd get better cooling but the bigger cell has proportionally smaller surface area, so physics would argue that the cells would need more cooling.
Another interesting theory is that they are going to switch to similar chemistry to what the A123 battery manufacturer used. It has higher cycle count than the standard Lithium Ion with Cobalt. It doesn't have Cobalt which is the most expensive part of the battery. They need more of the other materials, but the cost of extra materials will be offset by only using miniscule amounts of the most expensive material. But the A123 chemistry only has half the power density of Lithium Ion cells.
I guess the future will show us what is actually going on. Hindsight is 2020.