With past experience, GaAs parts’ switching speeds tend to increase by 100 to 200 ns at high temp.
For Silicon based parts, does the switching speed increase similarly as with GaAs parts?
I am not sure about which part you measured before, but I think it is really case dependent.
For ADRF5730, the typical switching speed is 125ns in the datasheet, and for HMC939ALP4E, it is 60ns typical. It is impossible for these two parts having switching speed increase by 100 to 200 ns at high temp (almost double or triple).
I actually checked the bench test data. For ADRF5730, the switching speed is increased by ~10ns at high temp. For HMC939ALP4E, it is increased by ~5ns at high temp.