There seem to be 3 approaches to LiFePo cells and battery packs
1. Ping (and others) pouches. Low C. Cheaper for high capacity. Best WHr to weight/volume ratio for LiFePo
2. Headway. Cylindrical cell. Medium C. 5,10,15,20 AHr per cell Now sold as packs with BMS by BMS Battery.
3. A123. Cylindrical cell. High C. Low capacity per cell (2.3AHr per cell). Only available from specialists like Cellman with BMS. Now turning up in all sorts of things like replacement M/C batteries for racing.
The trade off with LiFePo vs LiMn or the new composite LiNiCoMn is long life vs weight and price. The same AHr of battery is typically twice the weight and 50% more expensive. However they are also robust and seem easy to balance. The High C of A123 cells make them perhaps a better option than RC LiPo where high current is an issue. The low AHr per cell of A123 means you need an awful lot of them. eg 48v-10AHr is 64 cells
This makes me think that there's only two scenarios where A123 really make sense on E-Bikes.
1. DIY builds aiming for very high currents (more than 30A) as an alternative to RC Lipo
2. Very small packs for short scale commuting eg 36v-2.5Ahr or 5AHr. The High C rating means a 12s2p pack can support 25A which is at the high end of vaguely legal hub motor/controllers. A 12s1p (36v2.3AHr) is probably ok for the typical 250W legal set up.
If you expect to do a *lot* of miles, then perhaps the extra life of LiFePo makes sense. I had somebody suggest to me that the pace of change means that if you can get 3 years from the battery with moderate use, it's not really worth it since there'll be new options by then. Even in the year I've been following E-Bike tech, I've seen this happen.