So, I went shopping instead. I needed some bits and bobs to finish off the rear sets, and chased most of them down on the Internet. I was a little dubious about the left hand thread M6 x 1.0mm stainless steel nuts I bought off of E-Bay arriving in time as it did say "seller usually posts within 5 days", so I stopped off at my local supplier and bought some from them. The Rose joints, two right hand, two left hand thread, were ordered on the Thursday night, and arrived on Saturday, and the left hand M6 x 1.0mm die and die holder were ordered Friday afternoon and showed up on Saturday too. Sadly the other thing that showed up on Saturday, was a hangover.
While that lot won't prevent the rear brake from working, it will mean it doesn't work very well. For a little more effort, it ought to be possible to have a rear brake that performs very much like the factory one. My rear set levers have a ratio of 3:1, so I cut the original brake pedal up and welded a piece of 1"x 1/4" strap to it to turn it into a 1.66:1 idler for the brake mechanism. 3 times 1.66 comes to 4.98, and that's pretty close to the original 5.
As well as sorting out the pedal ratio, this also puts the brake rod in the original position where it's pivot is much closer to the swinging arms and the whole geometry of the set up is a lot happier and the rear brake set up now uses the original pedal stop. I might add that while the maths was getting more complicated, the hangover wasn't getting any better...
As you can imagine then it came as something of a relief to go back to making the link rod from the rear sets to the idler. Somewhere in the week, I did find the time to make the ends for the pedals, but apparently I didn't find time to swap the zinc plated Phillips screw holding the two sections of the pedal together for a stainless, button headed Allen bolt. I think it's fair to say that the finished thing looks quite a lot like a real one, though for the reasons I explained, they're going to work better than some of the off the shelf "real" ones you see.
Clearly, the original "bolt-on" concept I had for this project is long dead, but I'd like to think that the reasons for that aren't entirely due to my reluctance to part with money. If I feel I can make something that works and/or looks better, then I'd rather make it myself than buy it. That's not because of some misguided idea that you have to make it all yourself, it's simply because I know what I want, and I'm not compromising that by buying something that's only nearly what I want. The other part of the equation, which may come as surprise to people who watch a lot of Discovery Channel, is that it's better to hit some notional target of right, than it is to have it done in some notional period of time.