NavList:

Yes, this seems to be correct!

I have checked it using both the ex-meridian table and the cosine rule. It was a very good opportunity to learn—and immediately forget—the ex-meridian table! 😄

One thing to remember well: this time period is centered around noon—half of it is before, and half of it is after. The table gives 1′ or 0.1′ deviation from the maximum values at the boundaries of the time interval, respectively.

And the situation is even better than calculated here. For example, 1′ (moa) accuracy means that we should treat 1.4′ as 1′ according to the rules of rounding, so the time period is even longer for 1′ accuracy. Eyeballing the ex-meridian table gives about 14 minutes, and for 5″ (approximately 0.1 moa) accuracy it is about 4 minutes. However, we should not forget that the Sun is moving: its declination is changing at a maximum rate of 1′ per hour (around the solstices), which narrows this estimated period very slightly. In the 7-minute half-period in the previous example, this would cause about a 0.1′ error as a maximum, a rounding tolerance.