I don't, that is just a standard rigging formula. Like if you had a flagpole and you were using some wire rope to hold it up in a wind the effect of a 60 degree wire vs a 45 degree wire would be .5 to .7 (their Cos). At a zero angle, the line parallel to the pole, you get 0 and at 90 degrees to the top you get 1, full strength. Same deal with diamond wires on spars in boats.

These numbers do indicate the effect to the wider hubs is pretty minimal. I did a quick scale cad drawing using the inside diameter of my Ma2 26" rim, and using the full width of the hubs as the flange width for absence of the real numbers. There was 1.1 degree of difference in the angle, between a 145 and 135 and it added about 6.7%. Between a 135 and a 120 there was 11.6 %. Again, these figures are based on full width flanges, but they are proportionally correct, I hope. In the real world the makers may not proportion the flanges proportionally to the width.

The rim is the same as a flagpole in the example above except it is held firmly in a third dimesion, rather than being a strut going to the ground, the spokes don't know the difference. Spokes operate in three dimensions also, but for the purposes of the hub width I don't think it is relevant.

I am not an engineer so with luck someone can jump in and bash this stuff about a little. Though I did once model it for a wire on a boat using a quality spring scale, a weight and some cord and the results modelled very closely to the engineer's calcs. See, you can save money on college, though nobody hires you (with good reason :0)).