As far as I know, Tesla
still makes the Model S/X with 18650 cells, thus it's likely they will continue making the Model 3/Y with 21700 cells for some time. Throw in the power tool OEMs transitioning to 21700 and I'd expect the formfactor to see some additional development over the next few years much like the 18650.
I do not believe that the 46800 cell has manifest itself in a Tesla product yet. Given that 46mm D x 80mm L is larger than a D cell (33.2mm x 61.5mm), that's going to make for a rather chonky flashlight body relative to the 18650 lights that are rather common these days although two end-to-end would bear some resemblance to the 2D maglite formfactor that was common ~20 years ago. With ~5.4x the volume, one might expect those cells to clock in at 26Ah!
EDIT:
I ran new metrics using average values for the best 18650/21700/26650 density/power cells I'm aware of / could find:
I used values from
separate cells for peak Ah and peak I. The Synthetic C value, thus, is mostly for the sake of having something to work against on the back of this napkin.
My sense is that capacity (Ah) should scale in a mostly linear fashion assuming similarly efficient construction relative to existing cells.
Peak current for any possible high-performance version of this formfactor is another thing altogether - I doubt that a single cell will tolerate pushing >200A - especially when its surface area:volume ratio is but ~44% that of the 18650. Thus while the
naïve handwaved metric suggests 8.5C is possible, I imagine that <4C is more realistic while still providing
a heck of a lot of angry pixies.
Of course, Tesla has their own Secret Sauce™ for their specific cells and is apt to be interested almost exclusively in energy density since their designs throw massive parallelism at the problem which provides more than enough current to accelerate an EV at a satisfactory pace. Whether the rest of the industry picks up on the formfactor is very much an open question.