THE SHORT STORY:
I'm super happy with the results coming from these settings (the lens-sensor combo, including the Pixco speed booster)! What was super surprising was that these settings not only look good on the final files, but they also look good on the Panasonic GH4 flip-out screen (while you're shooting) and that doesn't always happen (it's a really good thing!)

THE LONG STORY:
I just tested this lens-sensor combination on a Panasonic GX85, so I had some expectations for this test...but I must say that it turned out a lot better than I expected! Not only does this turn the very affordable Canon FDn 50 1.8 into a great low light tool (a 35mm 1.2) but it renders quite a filmic look on this GH4 (a super affordable camera nowadays)! I say it looks a lot like film because it's doing a really good job controlling the highlight rolloff, midtones and shadow detail, and it even renders a sort of film grain look by having a bit of digital noise (but it's not too much that it's distracting). The NOISE REDUCTION settings are what I'm using to control how the "film grain" look is rendering, and the final settings I'm using are a result of a lot of back-and-forth trial and error, so I'll have to say that it's important to use these specific settings (try it and I think you'll like it).

In regard to the optical quality of this speed booster, it has flaws and imperfections, but I think what's really important is how the image it renders looks, etc. It has issues with lens flare which some people actually like, and has other issues that could normally be considered a negative (such as sharpness reduction, etc.) that are working well with this lens-sensor combination.

In the end, I think using this Pixco speed booster with this Canon FDn 50 1.8 is a very cost-effective way to turn this 50 1.8 into a 35mm F/1.2 low light tool. The aperture control is stepless (i.e. de-clicked) though it does have a couple places it seems to catch slightly.

GH4 NOTE:
One problem with the GH4 is that the sensor seems to render too much BLUE. So, the workaround I use is to shoot in "Cloudy" white balance when it's sunny outside. My "REAL" ratings rate "COLOR" as being pretty accurate, but it's when using this workaround.