No.
There is no point in that. The bottleneck is elsewhere - probably on the decoding source material in the timeline which done by the CPU.
The NVIDIA hardware encoder/decoder chip is the same no matter whether you buy a GTX1050, GTX1060, GTX1070, GTX1080 or higher.
The only speed difference would depend the clock frequency the board is running at. An overclocked GTX1050 can beat a GTX 1070 if we are solely looking at this function.
To test the real potential - take that source clip - convert it to DANIEL2 - then use the DANIEL2 encoded clip as source and to the H.264 or HEVC encode.
No you should see a much higher throughput. If not, then the bottleneck is somewhere else and your system needs further tuning.