If anyone wants to work out the pack sizes, then I can confirm that the cells in these packs are 135mm tall, 100mm wide and 4mm thick. The connections are on the top (the 135mm dimension) and add a few mm. The thickness of the cells is increased slightly in the packs by tape and the foam rubber used on the top, so if you use 5mm as the cell thickness it allows you to work out the thickness of the stack reasonably accurately.
Each cell is 4.95 Ah, 3.2V nominal so a 10Ah pack uses double cell modules, a 15Ah pack uses triple cell modules and a 20Ah pack uses quadruple cell modules. A double cell, 10Ah module measures 135mm x 100mm x 10mm, approximately.
The 12 way BMS circuit board is 115mm long, 70mm wide (including the connector) and about 10mm thick. A 48V board will be a bit longer I expect. Wrapped in it's heat shrink sleeve it measures maybe 12mm thick. This board will typically be taped to one end of the pack, so increasing one of the dimensions by about 10mm to 12mm.
The dimensions Li Ping quotes are conservative. He lists the 36V 10Ah pack as 150mm x 100mm x 150mm. In actual fact it measures about 140mm tall, 100mm wide and 145mm long overall, as it arrives, all wrapped in duct tape.
With a bit of guesswork you can pretty much predict the possible dimensions for any pack from the above.
On a separate topic, I put my pack through it's first long charge, as recommended in the instructions from Li Ping, and carefully monitored the voltage of each cell towards the end of the balancing process, during the phase when the charger is cutting in and out every minute or two. The charge process is interesting, as the balancing system causes the pack to slightly discharge, then recharge, quite a few times towards the end. I think that what the BMS does is to cut off the charge when a cell reaches the 3.66V point, then discharge the highest cell down to match the lowest cell, then restart the charge and repeat the process until all the cells are at 3.66V. The cell voltages were all spot-on 3.66 volts at the end of the charge process, so I'm reasonably certain that it's a pretty good BMS in terms of getting the most from the cells, and not a crude system as suggested elsewhere on here. I was seriously thinking of getting one of the excellent BMS boards from Gary Goodrum, but having monitored the way that the standard BMS works I think I'll stick with it, as it seems better than I expected it to be.
Overall, I think that these packs offer really excellent value, if we accept that they may not be as long-lasting or reliable as the more expensive batteries. Given that a few people on here seem to end up doing as I did with my NiMH packs, and having a incident due to "brain fade" that causes early battery failure, I'd rather do something daft with a cheap Chinese LiFePO4 pack than a more expensive, but better quality, pack.
Jeremy