Varactor voltage by thermistor voltage divider. Is it normal solution?

What if prescaler circuity gives too much precision, can it be replaced by thermistor voltage-divider before varactor? I think curve can be approximated by thermistor. Actually i worry about active device, maybe there some "aging" of it capacitance, etc., so thermistor will give wrong correction after some time... when device change it frequency curve over temperature.

I wonder if it might be possible to phrase an understandable question from this line up of terms?

does active device parasitics, capacitance, etc. change over time? all that things that can influence frequency. p.s. somebody sure can phrase a whole patent in such terms

You can surely use thermistor but tuning voltage is almost constant and it does not create a trouble.
The main effects come from other temperature sensitive components and they can not be easily tolerated and compansated.
For instance, temperature dependency of a intrinsic capacitor can not be compansated in any manner, in this case what does it make sense to tolerate tuning voltage by using thermistor ?
Nope..
Frequency stability of a VCO can be handled by a simply well designed PLL.