Thanks Oddjob. I was just looking at the paint for further confirmation really, had no intention of using it for replacements. As you rightly say some of it has deteriorated, but some is still surprisingly clear.
I would rush to point out that I am no mechanic and have no experience with Honda four rebuilds at all. One week ago I had no idea about the colour codes on Honda's shells, but Oddjobs superb explanation sorted that out.
The real problem started when I could not really be sure about 2 of the numbers etched onto the crank webs. The first two could have been either 1 or 2, so being the engineer I wasn't settled with that. Then for No.5 the black shells don't exist any more. Then the actual Honda workshop manual does not even mention crank numbers, it advises to measure yourself. I realise that most people don't have the tools to measure a crank to a 1/10th thou, but with my previous job as a Toolmaker, I do (your local engineering shop would measure it for you cheaply). And when measured it revealed that the crank etchings on two of the webs were in fact wrong. If any member that lives near me a Blackpool would like to pop in (kettles on) and re-check it they are most welcome. I am 100% sure the original measurements contain mistakes.
On No.5 it was fitted with a tighter shell (black), and the amount of wear showing on that shell is more than on the others. The thing is though, the difference between the largest diameter case bore to the smallest crank journal is only 1.8 thou, not a lot. You wouldn't want to, but you could probably fit any shell to any journal and with careful running in it would be alright. That's possibly why you can only get Green and Brown now, the middle tolerance bearings.
For Honda to make a crank to such good tolerances is nothing short of superb, but they have Human Beings in the mix and that means mistakes. It is just impossible that Honda turned out every engine to perfection with zero errors. People make mistakes, that's why they put rubbers on the end of pencils.
By the way talking of mistakes, the Haynes manual has a mistake in the shell sizing table, misprinted tolerances.
I'm happy with my measurements, and all you can do is go with what the actual sizes are. If it seizes up first time out I shall fall on my sword.
Big thanks to Oddjob for a top explanation of how the coding system works, cheer mate!!
PS. Just looked, and this has taken pages of the thread. Sorry, Just hope it helps someone else.