Martyn,
This was one of the inspirations for my original post, in that the figures are given without a discussion of how they were arrived at. It's said that modern fuel is the relevant factor, suggesting it burns faster than when the factory specified 39 degrees, but no one here seems to know why or to what extent (despite a great deal of informative technical discussion on ignition advance).
Assuming a standard Vincent twin, with compression ratio in the range the factory used (6.7 - 8:1) and original magneto/ATD, fuel composition is the only factor I can think of that would invalidate the factory's original specification. To see if it does, we could look at fuel burn characteristics (seemingly unavailable) or dyno tests on actual standard bikes running various fuels (race gas, pump unleaded, maybe some leaded premium if available) - but it seems no one has done this either.
Lacking hard data, we can fall back on "it seems to run best at x degrees" - but as davidd points out, seat-of-pants impressions are notoriously unreliable.
In my case figures for maximum ignition advance were arrived at with a combination of on-road testing, feedback from a number of Vincent riders (you would be surprised at just how many Vincent owners don't ride their Garage Queens) and observing the effect of too much advance.
The advance recommendation at maximum retard for starting - 4 BTDC - is the factory number and has been repeated a number of times in publications written by Phil Irving - the designer of our engines.
I had an original Lucas ATD fail when out touring at speed. It was early in my 'vincent' life and the ATD had been tuned by bending the stop ears on it. Well unbeknown to me BOTH the stop ears eventually broke away and I subsequently found - after the pull down that it was providing 4 BTDC to start and 46 BTCD fully advanced. I did notice some pinking noise on the hills but cruising at 75-80 mph it did not seem too bad. Here was the result it had on a CP Forged piston, and yes suddenly I lost a bit of power.
As you can see from the carbon deposits - it was not running lean - in fact it was (intentionally) a wisker rich. New piston, bore cleaned up and a new ATD was the main work needed.
Modern motors are fitted with knock sensors and while i'm no expert I figure that they work by retarding the ignition if knocking is detected. Knocking or pinking is the noises the motor produces when the fuel charge burn reaches maximum pressure BEFORE the piston gets to TDC and in doing so tries to force the piston to go back down the bore BEFORE it is at or past TDC. Produces noise, excessive heat on the piston crown and also beats up on the little and big end bearings.
The only knock or pinking detectors we have with our old bikes is our own senses - mainly hearing. If at any time you hear those sounds then you know the charge is being ignited/burnt too early and the only way to stop it is by 1. closing the throttle so there is no fuel to get burnt or 2. ensure that the charge gets burnt a bit later - ie by retarding the ignition. And that was the basis of my recommendations after lots of road testing by myself and a lot of other Vincent owners.
Purists may suggest I have been too conservative however the consequences of too much advance is in my experience catastrophic and expensive.
But then again as David D has earlier suggested (post 24) , the ONLY definitive method is to use a dyno to determine the setting for maximum performance which by the way may well NOT equate to maximum reliability.
Martyn