Rare has no inherent strict meaning; only when defined by an authority as in John's instance of pocket watches and even then, only when that definition is accepted by a wider group of people.

So what John is doing is applying a norm he learned in a different, albeit tangentially related, field. Which is fine for him but not something we have to accept universally, although we may choose to as above.

Personally, I'm not keen on an absolute definition of rare that allows enumeration and prefer a comparative measure.

Take the 1153 BN that Jeff used in his grail article for instance. A very nice watch indeed and hard to find in ideal condition but there is also an SN that is at least one order of magnitude scarcer, possibly two. Is it as nice a watch as the BN? Arguably not, but it is definitely rarer.

And not just rarer, it would actually meet my definition of rare - in 25 years I have seen maybe four. But I am also sure that more than John's threshold for rare were made.

What I'm getting at is that John and I appear to have different definitions of rare and that that's absolutely fine. It becomes more of an issue in an auction environment when terms like that are used without necessarily having a shared understanding - the best approach would be for the auction house to have a glossary elaborating its specific use of general terms. How many auction houses have you seen that were that diligent?