Since you're using CFR with x64
You mean CRF (constant rate factor) with x264
CMoar: CRF effectively means the quality will be constant, which means the bitrate will not be constant and the filesize is unpredictable. The obvious factor affecting the filesize is the running time of the video. Then the complexity of the content: firstly detail, e.g. trees or bushes - lots of bits needed to store that compared to say, a plain white wall; also rapid/unpredictable movement, e.g. a hand-held camera, or those same leaves blowing in the wind etc. So generally, indoor shoots compress well, and outdoor shoots don't (especially in natural environments).
So you need to judge the complexity+length of the video before ripping, and from that make a decision about CRF. If a video is long/complex then it's often compressed badly already (if the DVDISO is still around 4GB). There's no point storing it with excessive bitrate if the detail/quality isn't there to start with. I usually adjust my CRFs up or down depending on complexity and how much I like the video. Only experience will tell you how much to adjust (but if you add 3.0 then the filesize halves, subtract 3.0 and it doubles, so the adjustments are usually much less than that).
___
You're pushing too much, 64 hours is not worthwhile unless there's something special about the rip (e.g. I'm currently repairing a BDISO (Rei) that has serious bitrate problems, the detail pulses on and off. I'm running a six-pass process to repair it, probably the most complex script I've ever used. Even that will take < 60 hours at 1080p@60fps). The fact you can't get anywhere near 100% CPU is your problem, and really the only way you're going hit 100% CPU with those settings is to learn to rip manually so you can multi-process + multi-thread (which is how I can run insane settings without dying of old age first).
Are you comparing the output of your rips to see whether your settings are giving worthwhile improvements? I provided the trim settings so you could make test rips quickly. Every video is different, some don't benefit from source match, others need more smoothing or less noise etc. Ripping at this level needs to be methodical or it's time wasted, there are no perfect settings that are best for everything. It will also help you see how little you are getting for those extra 60 hours of rip time (diminishing returns...)
BTW. You will never hit utmost quality. I mentioned the most accurate settings earlier, but in reality there are better settings/scripts than that. They're just so slow that no-one could ever use them. Firstly, QTGMC works better if you run it twice, with the second pass using the output of the first to guide it (1st pass Placebo, 2nd pass Very Slow). Secondly, there are two source match presets that I don't expose in MeguIVit. By default both settings are around Very Fast, yet they both go to Placebo. There are complex pre-filters for motion analysis, high quality noise processing, EEDI3+NNEDI3 combos for interpolation, etc. etc. etc. Would take a year for a rip if you used it all. Once you accept you cannot hit ultimate quality then you can feel OK when you step it back to something more sensible...
