This video is extremely tricky to encode due to a few reasons:
*visible noise in the Y channel even when downsampled.
*Lots of motion over a single shot in a short ammount of time
*smoke has a lot of gradients, encoders will try to crush this information because of their default matrix of where bits are spent.
*short clip, so a 2 pass encode doesn't give you a lot of room to do a spike in bitrate.
No setting that you change in Compressor will give you a better result. You will need to pre-process your footage before you hand it to your encoder to get a sharper image from your example.