Live View (at 720h) has better res than the old analogue cameras, where peaking originated. On those (going back 35-40 years!), it was really easy to use, and, IIRC when colour became ubiquitous, usually on the green channel only. At least, cameramen had the option of selecting which channel to use in the (monochrome) viewfinder, and green was favourite*. Studios were lit, IIRC, to give working apertures around f/5.6 at 0dB gain (i.e. normal camera channel lineup).
My point: peaking seemed to work pretty well back then, even though the cameras were lower bandwidth than today (in analogue terms), and there was significant noise in the channel. It was a substitute for being able to see sufficient detail in the camera's viewfinder, and even in well-lit sets, it wasn't unknown for the lighting director to hit production talkback and yell "focus!".
I guess with an HD source available you might arguably filter it for "higher frequency", but I can't see how you gain anything in practice: if you're using a fast lens wide open, you'll have a narrow DoF, but if stopped down you'll have more, irrespective of the resolution you're working to.
So it doesn't matter: peaking "at lower frequencies" will still show you where the lens is focused. In any case, with a wide open lens, you'd either have to rehearse well, or risk the camera operator hunting through focus during the take.
I've never been a D.P.: It may be they want something different on the monitor, but surely that's a different issue to the original purpose of viewfinder peaking -- to be an aid for the camera operator.
S (ex- BBC Bristol, but an Audio Supervisor, not a camera person!).
PS: I mainly do stills (6D): I find the present system extraordinarily useful and use the Live View version.
*In the UK, from the late 1960s through to the start of the 1990s, we used one odd but very popular camera, the EMI 2001 (they were sold elsewhere too). It had a fourth tube for the luminance signal. This was an engineering kludge - the blue channel was very low output for a number of reasons, and considered too noisy to just combine the three channels to derive the luminance signal arithmetically. I can't remember, but at a guess the camera's v/f could probably be switched to use luminance, as it would have been a lot higher output than the green channel alone. If anyone cares, I know who to ask!