I don't think you can really do a replica. Elevation changes are one thing, geometry of the corners another, so to do it properly it has to be routed (or cut up lots of sectional track) and permanent. Even then I doubt it would be close to scale.
And once you have done all that, is there any guarantee what you end up with will make for a good slot car track?
Having said that, if a person decides to build a track based on something real, then why not? It is not carved in stone that it won't result in a good slot car track. I think the key there is to pick a track that appeals on the grounds of having the things a challenging slot car track also needs: non-constant radius turns, a good mixture of fast, flowy and technical bits and one long straight (for example).
The best HO track I have driven on is the Derby 24hr track, followed closely by Pinewood because of the faster cars used. One is a real track, a pretty good flat copy thanks to an American by the name of Brad Bowman (same chap who makes our BSTS in fact), although not to scale of course, the other totally fictitious. I doubt the fact the 24hr track is 'real' is what makes it so good, in fact for me (being a LeMans ignorati) I know it is not.
The other regular contenders one hears talked about as best HO track is the KSR (Bowman again), which is not a real track, rather inspired by several, and LeMonzaCo (made up of the best bits of real tracks as the name suggests, but not of itself 'real').
As I said to you at the time, what you have planned for August looks cracking on paper, only you can decide if you want to stick with that way of designing tracks. I think we both feel building real tracks scaled down is not automatically the be all and end all, right?