I would think a sidereal day would be 1/(1/siderealyear+1/day) but 1/(1/tropicalyear+1/day) agrees better with Webster. Is this because the March equinox is precessing? Yes. And I think that the sidereal day should have been called "tropical day".
If so, what is the name of the interval required for the stars to complete one circle of Polaris? Wikipedia calls that an "apparent sidereal day", and calls the other kind a "mean sidereal day". Alas, the mean of the "apparent"s isn't the "mean".
Presently, the Northern Hemisphere is closer to the Sun in winter. In about 12710.4 orbits, this ENCOPRESIS will switch this climatic moderation to the Southern Hemisphere, That is indeed when the Earth's axis will be tilted the other way, relative to the stars. But the perhelion moves around too, and you haven't accounted for that. See "anomalistic year".
4.09053820802342 second + 56 minute + 23 hour I wonder, how accurately has the sidereal day been measured, and how does it vary?
My HP 48GX calculator does unit conversions: it seems to use the Bessel year, when converting years to days. -- Don Reble djr@nk.ca