There is a limit as to how much you can reduce the "temperature sensitivity" of a load. It's not just the ambient temperature effect on the loaded rounds, it's also the temperature of the barrel/chamber. Brass is a very good conductor and even chilled rounds will warm up very quickly in a warm chamber. Further, a significant amount of the powder burn will take place within the barrel. A hot barrel/chamber can also increase velocity via a temperature-dependent effect. To truly test the effects of temperature, the whole system needs to be equilibrated at some specific temperature, not just the loaded rounds.
If you're looking for some load that will show minimal velocity change over a 50-60 degree ambient temperature range, forget it, it won't happen with any powder ever made. If you're simply looking for a load you can shoot a fair bit of the year with good precision, I would optimize it such that you're not going over pressure during the warmer months, let's say an upper limit of 85-95 degrees. In this case, when the temperature drops you're not going to have to worry about blown primers and/or other pressure issues. This will not be true if optimizing a load for colder average temps, so I'd rather not have to worry about over pressure issues...much better to err on the side of caution. With a load optimized for the warmer months, you might experience some drop off in precision when the temp dips 20-30 degrees (or more), but it shouldn't be huge.
If you really want to optimize a load to normalize temperature effects during the season, you need to record good temperature/velocity data over a range of different temperatures for specific charge weights with a given Lot of powder. A program such as Quickload will give you temperature-specific velocity predictions, but how accurate they will be for your specific setup can only be determined by empirical testing. Once you have some idea of a) how much velocity change you observe per degree F; and b) how much velocity change you observe per 0.1 gr charge weight, you can make some fairly good estimates of how much you need to adjust your charge weight to keep velocity within a certain window, provided you're not rapid firing and heating the barrel up dramatically during a string of fire. It's really not that difficult to make minor changes in charge weight during the course of the seasons, I do it regularly. I don't worry about seating depth for this approach, just minor changes in charge weight. But you have to have solid velocity data as the basis for making the correct changes. For your specific purpose, Varget or H4895 might be good choices; both are fairly resistant to temperature-dependent change in burn rate as compared to some other powder choices. However, even these two will show temp-dependent effects; i.e they're not totally resistant, which is why adjustment of charge weight is the easiest approach.