@rust, we already know the true average speed from the timetables. What is not known is how much time the buses spend stationary and how fast they travel when moving.
Lothian Buses are basically arguing that while service 44 averages 12mph (they admit this, it's their published timetable) that when the buses are moving they go much faster than 20mph - so if a 20 zone was put in place it would compromise the service.
Unfortunately the bus tracker app can't help with this - we'd need the actual bus GPS tracks (assuming that's how the system works). It's tempting to submit a FOI but I'm guessing they'd weasel out of it under 'not reasonable' or 'too time consuming'.
The only real way is for upstanding local heroes (hi, rosscbrown!) to record GPS traces of actual journeys. It's then trivial to analyse the number of seconds spent travelling at any given speed, diff the area under the graph with one driven at a max of 20mph to get total delay.
Even then it's ultra conservative, my original article assumed that each second translated to a second's delay whereas of course it should be modified by the proportion of time the buses spend stuck in queues anyway...