What follows is only my personal opinion of course 🙂
Re the current Zimmerman algorithm I’d like to be able to change the % factors (4 per Farhe and 1 per deg mean hum) and also the thresholds (70F for temp and 30% for hum). I live in a coastal town and the current computed values are way too low since the average hum is always VERY high.
Also empirically I believe the average (?) wind speed could be usefully benefit in the simple algorithm since windy days tend to dry the plants.
ET would be far more sophisticated probably. But the simpler algo with modifiable parameters would be very good I think.
Are you looking at this ? http://en.wikipedia.org/wiki/Penman–Monteith_equation
The only problem I see is the relative scarcity of meteostations which give radiation values (my neighbour’s station I use has it luckily) and the fact that for instance a lot of my little garden gets the shadow of buildings and one large tree and therefore the radiance factor in ET would be almost totally useless.
Thanks so much for making OS more and more of a GREAT BEAUTIFUL USEFUL tool !