Even with alkaline cells there is not that much energy left at low voltage. If one would want to use a alkaline cell to a lower voltage one may need 1 more cell. In the current example 4 cells would allow 0.9 V per cell to get the same final 3.6 drop out point, but one kind of wastes the energy from 1 more cell. So the 1.2 V final voltage is not such a bad choice.
With a linear regulator and a fixed min voltage getting the ideal final voltage is a little similar to the maximum power point of a solar cell. Depending on the voltage chosen one gets a different capacity / current and there is a point of maximum power / energy.
Going at least down to 1.2 V allows to use NiMH or NiCd cell.
This gets very anoying when the drop out is higher (e.g. 1.4 V in some cases
![Sad :(](https://www.eevblog.com/forum/Smileys/default/xsad.gif.pagespeed.ic.L3FGyzQrjB.png)
).
Extrapolating to lower current draw is relatively easy, as the low currents the batteries are usually well bahaved. So there is a capacity for low currents and than a drop at the higher current end.
So for the 5 mA load using 2 times the time for 10 mA load is a good approximation. So no need to really extrapolate the curve.