Ok, i must have misundertood your original post.
For example… For me to use bezier’s, I merely have to click a few times with my workspace saved as it is, and in the context of colour correction I create masks just easy as the greatest programs. So vignettes and masking are easy for me. The speed of creating a power window in blender for me is not much slower (although it is slower - by a few seconds I guess) than the process of doing one in Color.
I have to keep reminding myself that you are talking about Colour Correction / Grading only, as a seperate process to the whole NLE thing. Everything you are talking about is in nodes and working on whole video, not a live edit, is the terminology an ‘Offline’? not in the VSE. Or do you have some secret dynamic link capability with FCP :-)?
Goes back to the idea of being able to apply any number of node groups to any number of VSE strips.
For small quick projects I’d assume that for speed and efficiency the ability to CC a live edit is the most practical way to go or is it a no no, should I be doing an intermediate non CC’d master in .png’s or lossless codec and re import to CC seperately, then final master, then encode?
Obviously for big movies and such the CC goes to a seperate department or studio, a DI? and an intermediate format is used. I guess emulating that process would relief some pain. 
I do it all in the same process, import, edit, CC and find it a pain in blender, but enjoyable. 
But definetely going to try the import of a master and CC as a wholely seperate process as you yourself are doing but don’t know how I’m going to see the scopes/histogram for feedback other than using a scene on the VSE. 
btw did you sort out your intermediate video format from your other post?
True. I would very much like it if we didn’t have to click on ‘Refresh’ every time we had to see the changes made to the Vectorscope and Luma Waveform monitors.
Is this to see an update of a node scene on the VSE? I don’t have to hit Refresh when working in the VSE? EDIT Read further down and see you mean the UV Image window. 
Not for me at all. Read ‘The Art and Technique of Digital Color Correction’. It is written by a professional colourist that throughout the book sits with other colourists that work on many blockbuster and independent films as well as TV shows, and throughout the book him and other colourists demonstrate the power of using curves instead of the Lift,Gamma,Gain sliders. They use curves for Saturation and Hue adjustment. They are in Color’s Primary In Room. They are in Avid Symphony Curves Tools. They are in Avid Xpress Pro Curve Tools. They are in Color Finesse’s Curves Tab. And it is all very easy to set up in blender.
Yes I own a copy of Steve Hullfish & Jamie Fowlers book and read it many times, aaapting the tutorials to blender including use of curves, but there are no scopes, histogram or zebra nodes to check on progress. So if the node scene is on the VSE to do that then I would have to reafirm my comment of the disfunctional connection between the two. But this goes back to my comment about offline. I do most in the VSE and see that as the weak area, nodes is ok for offline, but a pain for live edit. Perhaps CC is not a live edit process?
The push pull technique is so simple in blender, I always start with 4 RGB curves for colour correction. Just as these colourists in the book do. The first I adjust the black and white points. The rest I can play with the colours etc, in their different areas - shadows, mids, highlights. In that book the guys pull the Red and Green curves down to create a dark blue in the shadows, and for the red in the highlights they push the Red up towards the highlights, for the gamma you can push it wherever. You can dial down the saturation easily. And preview immediately in the UV/Image editor. …Though the ‘Refresh’ button has to be pressed to see the graph changes… grrr…
How do you keep a check on you the frame you’re working on without zebras, RGB Parade and Luma scope and histogram. Is it all by eye? Perhaps I’m over using them and need better trained eyes. 
As I mentioned earlier. I don’t find it painful to do so in blender. Though I guess it is a little slower. Than some of the programs. Though I don’t find Magic Bullet Looks/Colorista to exceed blender in this regard necessarily…for example in the clip, the guy himself states that Colorista can only do a square and an oval, though he says anything else is never used anyway, I disagree: (an excerpt from my book)
"One of the most adamant users of these custom shapes for sec-
ondary color correction is veteran Chicago colorist Pete Jannotta: “I
always prefer to draw (custom shapes) because even in da Vinci I don’t
use a fixed circle ever anymore. What I like to do is draw the shape
and then move it around so I can see how the light is working. I
don’t care what the shape is, but if it’s an oval, then it’s fixed and I
can’t control it.”
Many, many, many of the colorists go on to say that they use custom shapes, because they are sick of using the typical same shapes. In fact, from what I remember less of the colourists out of the bunch that are doing corrections in the book actually use the square and oval.
Stu (The Spirit) also did the RebelCC stuff for After Effects which is the tool he uses most and in his book Rebel DV he uses beziers for so called power windows, so I guess he was in selling MB/Colorista mode. 
And on another point the guy in the video speaks of the fact that Magic Bullet Looks cannot animate vignettes! That is not necessarily better in any way, I figure if blender can fill the hole that both of these have without being used together, then blender is not necessarily falling behind in the vignette area and if it is then it is not falling behind very far.
Yes a limitation, he does show that Colorista can do it. It’s swings and roundabouts as they say. 
I’ve seen that video before and I love it. Everything in that video I’ve done in blender quite quickly… once again I’ve changed my workspace to fit it! Though blender isn’t at their level in terms of appears or user friendliness, or default speed. If one does not have the money to purchase Color or Da Vinci, I think they can rest comfortably in the fact that blender can handle those tasks well… also if they do end up going to work somewhere as a colourist, it does not take alot to get used to another film colouring system…it’s very easy to bounce between different programs and pick them up immediately.
Then we agree. 
I guess just want more blender colourists to emerge I guess. It’s not like Maya or 3ds Max, where the industry needs you to be able to adjust quickly to their programs for professional work. The underlying principles in colour correction are far more easily transferable between programs.
Perhaps it would be useful to get some tutorials and .blends together to show blenders abilities in this area.
I adore Harry Potter! The books, the look of the film I just die over!
Do you have any techniques for things like faking shallow depth of field, diffusion etc. I know a lot of that is best in camera but it can be hard work with consumer/prosumer level video without getting into lens adaptors 