I thought I had worked out the optimal relationships to get these curves looking good. Turns out... no! It's been a long time since I did anything on a huge scale and now I remember why -- what a pain. I probably could get smoother curves if I had the technical chops, but for better or worse I don't have a mathematical brain at all.
As for a believable upper case: as can be seen I'm in two minds regards diagonals in the lower case; this will be compounded in any upper case scheme.
Nevertheless I wanted to share this unfinishable thing because I still find the forms pleasing, unresolved though they may be.

Feel free! It would be interesting to see someone else's take on it.

Unfortunately I have a recurring illness that knocks me out once every month or so for a week. It's tedious, but I'm learning to cope with it. But all the time I've been quizzing myself how I'm going to believably extrapolate a Cap O from what I've already got. So I do intend to get back to it when I have the stamina for it.
I figure I'm going to have to take a screen capture of the lower case o and bring it into InDesign and blow it up, to make sure I get the right proportions.
I had misgivings about the v but seeing how you like it I'll keep it! The x is a problem, though, don't you think? I'm not sold on either options I've come up with.

This is doing my head in. How to go about increasing the radius/weight for the cap version. I've also noticed to my dismay I've made a mistake with the placement of some characters on the baseline. That'll serve me for idiotically using the "flip vertical" function without engaging the brain first.

It's a mug's game trying to do this sort of thing. You can only increase the radius of the upper case bowl by so much -- and it rather governs the width of certain letters. It might look regular but it's full of vagaries when you look up close. Oh well.

@Goatmeal: Ha ha! We could do with Mr Meek giving us a few extra guides so we can easily keep track of things like cap height, l/c ascenders and descenders, x-height, bowl overshoots and the like. Preferably a different colour for each. (Now I'm being just silly.)
@sarreyn: ain't that the truth. After creating a roughly-spaced-out u/c O I spent a whole week trying to work up the courage to tackle the issue of increased radius. As you say, them limited angles. There are possibly more composite permutations than I've been able to discover; I've not been particularly methodical about investigating them. It's like searching for the lost chord: I still find myself trying to re-invent the wheel. "Ah, I know! If I combine those bricks into this composite... nah, you've already tried that, sunshine."
I'm reminded of the early days of imageset typesetting. Glyphs bigger than 72pt revealed the same, ropey combinations of angles to approximate curves. More sophisticated because mathematically derived and optimised, but the same process. Now I'm really showing my age.

It's easy to see I haven't been reading the updates! Thanks for the tip.