i always guestimate that the half way point of the cylinder is around the half way point of the two temperatures
so at half way i'd guess your water was about 33C
is this logical? i suppose it must be close
Well, actually no. Suppose you start with a tank full (top to bottom equal temp) of hot, and then draw off some HW. The bottom cools faster then the top, and there is no way you can know from only two sensors what the profile over the entire height is.
I've got three sensors, and the middle one very rarely is near the average temp of top and bottom.
That said, you've got no choice. Short of installing more sensors your averaging is probably as good as any method. Just don't think that the calculated energy values are 100% correct

----
Forecast for 1 August doesn't look promissing, does it?
Klaus