This works as a rough estimate however some manufacturers list the AFOV larger than it really is as well as rounding off the F/L which can also vary the true FOV some.
it is a rough estimate but I doubt the afov cheating is enough to change the basic structure of the comparison. As for example 2788 vs 2800. You'd have to get out the calipers to see which one really offered more field of view; and in the end, they are both designed to offer maximum field in two inch format, even if the 2788 "won" (more field than the 2800) it would not be by enough to matter.
There are other parameters that also change. For example my f/6.9 CFF clocks in, to be precise, at f/6.85, and my old 10" f/6 Newt was, in fact, f/6.3. And one will find similar variations around the stated specs in other brands. So if you have two Stowaways one might be f/6.66 and another one f/6.7, which will affect true field between two different copies.
When I have clocked star transits at the celestial equator (transits across the field of view) I generally get results within 5% of what you would expect from the specs on the gear. Some of it is measurement error: it's hard to have a star enter the exact center of the ocular in such away as to cross the exact center of the field of view. And there's some delay in clicking the stop watch, and it's hard sometimes to say *exactly* when a star has left the field. Because there is an ever so slightly vignetted edge where the star is part visible part not. And then there's the variation in scope focal lengths, and the variations in the oculars.
Taken all in all 95% accuracy is probably pretty good, and in a context where that is the case, the calculation trick I outlined above is well within the norms of the gear we're using, taking them as total systems.