Hybrid has been using a lazy DXVA based HW decoder, it must be change

[HW decoder]

Selur, why don't you integrate more free frame accurate HW decoder engines and options to the Hybrid which are faster than the present DXVA based decoder?

a. because nobody asked for it. I added support for selecting the hwaccell decoder used in ffmpeg a few days ago, see: https://forum.selur.net/showthread.php?tid=494 Note that this will only help if ffmpeg is the decoder of the source material.
So don't use Avisynth or Vapoursynth filters and no DVD input.
b. Hybrid does support DGDevNV (Avisynth), DGDecIM (Avisynth&Vapoursynth) and FRIM (Avisynth) as hardware based decoders.
c. when I implemented the decoder support ffmpeg only supported DXVA on Windows and I and for me software decoding is most times faster, since I rarely encode stuff which benefits from gpu support which my current gpu support.

Cu Selur

DGDevNV is not a very advanced decoder and slow. Why don't you implement NVENC decoder, or other more efficient Cuda decoders? (Which can use the real speed of modern CUDA based Nvidia cards... 

I can't see any selector between hw decoders in Hybrid. Maybe you can add a similar simple quick selector button like StaxRip has. You can put it into a drop-down-menu in the the config =>input panel, instead of the simple "use GPU decoding" option.

There are huge differences and impact of decoders during the transcoding process. With the same encoder settings in the Staxrip, a slow HW decoder can do 50-80FPS, a faster HW decoder can feed the encoding process so fast, that it is over 400 - 500 FPS. That's why the selection option for decoders are so important. So it has very, very important effect, especially if you compress long blu-ray films.

Edit: DG (Donald Graft ???) decoders are really terrible slow, and I can see on my PC and Laptop, that they use the GPU extreme minimal level, unlike other decoders which use them much more, thus they produce a very fast feed for encoder process.