Sorry to get into this late,
but this forum's format is unaccessible to my older PCs.
It may seem counter-intuitive for rich-running to suggest a vacuum leak, but it can easily be the result of over-enrichening to accomodate a leak that peaks at a particular rpm.
Even with modern compliant manifold seals, it is still crucial to bubbletest the assembly, as leaks can occur at other interfaces than just the seals. Casting porosities and other unforseen leaks can then be detected, independent of rpm, temperature, etc: http://virtualindian.org/11techleaktest.html
There can be condition concerns within the carb of course, but until the possibility of even the tiniest of vacuum leaks are absolutely
eliminated, tuning will always be a tail-chase.
With zero leaks, even a worn Linkert is remarkably forgiving!
Let us know how it goes,