First folks should understand this is not a 4 channel firmware. It's actually many different firmwares programmable by the c-preprocessor options, and 4 channel is just one option. You can build it for only one channel or two or three, and you can select whichever of the 4 channels you want (although 1 and 2, on pins pb0 and pb1 are preferred slightly at least). 4-channel was really just a why not kind of thing after figuring out how it was possible. The Vcc-read and the eswitch ire probably the most immediately useful features on existing hardware and the OTSM will I hope be a great benefit in the very future possibly even on existing boards with some changes to the components. It also combines well with the other two features.
Color mixing was pretty much what I thought you were getting at David. But before someone can program it they'd need a vision of how it would be used. If you just want to set up some color for photography I guess it can require a little setup and several menus for each led to fine tune things. Using a standard UI I'd probably do color mixing at a medium brightness level in the hidden modes or just have a dedicated modegroup like that, just have as many preset mixes as you want. Then standard modes are just W or maybe equal RGBW, or anyway a fixed custom tint.
One thing I've thought would be nice though, similar to hidden modes, but somehow make it easy to switch back and forth between two or three mode groups. Not the full menu 5 thing, some kind of quick access toggle. It could be a short menu in the hidden modes. Then the user can define their 2 or three bookmarked groups from the main menu. For me I use my light walking and biking and indoors, and kind of want different mode groups for all three (but especially for indoor vs outdoor). I wouldn't mind if it even defaulted back to the indoor settings after an hour. (RAM actually lasts that long with these big caps, but without an eswitch there's no good timer and the RAM probably actually lasts too long).
Anyway, lots for room for creative advances still. I was kind of hoping to mostly knock out the real hardware-driver side of the firmware with as many configurations as can cram into 5 pins. Then the "application" side of the software can be developed more, but I'm sure there will be more ideas on the hardware side too. There already are a couple of ideas that fall in between, like pseudo regulated PWM, that aren't in here. If I do more that might be my next thing.