From my experience, lowering the temperature generally reduces the activity of the Gal4-UAS system, even without the presence of Gal80TS.

The best temperature for the expression and activity of GAL4 is 25C, so lower temperature reduces its activity. I also observed this during my research time working with GAL4.

Yes, the expression of Gal4/UAS is temperature sensitive (optimum is theoretically at 29).

As per the previous responses the efficacy of GAL4-UAS expression levels can be temperature sensitive, with 25ºC generally seen as the median optimum temperature. You can often observe variant levels of phenotypic penetrance as the temperature the is altered; depending on what the constructs are/what gene expression they impinge this can in fact be quantified (ie. survival/viability assays, levels of fluorescent expression for reporters, fertility, behavioural outputs etc.). Generally RNAi is said to work best at 29ºC, though again this may come at a cost as this temperature can heat shock the fly, affect fertility etc. If you have lethality occurring at a higher temp (ie. 29ºC) you can sometimes get 'escapers' if you lower the temperature to 25ºC down to 18ºC. Note - If you are trying to quantify expression levels (ie. qPCR) or do comparisons you really need to ensure the parental controls are also raised at the same temp/conditions if at all possible.

But yes, if you raise larvae/pupae at 18ºC there is a good chance you will reduce efficacy of GAL4/UAS expression. If your GAL4 driver/Gene of Interest is expressed during these points of development this can alter later effects within the developing animals. FlyAtlas is a good point of call for determining thgis (https://flyatlas.gla.ac.uk/FlyAtlas2/index.html?page=gene), but you could also isolate RNA from developing animals at differing stages and differing conditions to also gain insight into possible levels of expression?