I knew the guy who made this
He was in a fairly flat area, so did not need a lot of hill climbing torque. I worked out the torque needed to propel a driver & passenger up a 20% hill + light vehicle weight, & an allowance for a small amount of luggage. I know more amps = more torque, while more volts = higher speed. Given that the hill climb may be low speed (same effect on ICE vehicle) I wondered if putting the batts in parallel to climb the hill, then switching to series to get the cruising speed on the flat could give economies of both cost & weight. As trex says two 36v 30A batts in series for the speed part, then going to parallel for the hill climb. Two things, 1) that LVC for either batt would not need to be a function of the controller, 2) the controller would need to be happy working at both 36v & 72v, with the heavy amps capacity needed when on 36v.
I believe it is not practical to use any current geared motor in this way, Xlyte have a 600 series motor aimed at motor cycle conversion. The direct motors being able to take overvolting without failures.
I doubt this could presently take safely to the UK's congested roads. Battery cost is a huge factor in any vehicle beyond a conventional bicycle. See the idea as a motorbike & sidecar in the 1930s (the original introduction to powered transport for the masses) vs the cost of say a Nissan Leaf