Sorry for the late reply here. The reason why I used 2% instead of 5.6% is that we’re doing a 3D View analysis here (not a 2D one) and I took 20 degrees to be one dimension of the the solid angle of the view. So I interpret a “quality view” to be one that is 20 degrees by 20 degrees OR a non-square view with an solid angle equivalent to 20 deg x 20 deg.
2% makes sense when you consider that, when _viewTypeOrPoints is set to 1 - Horizontal 60 Degree Cone of Vision on the “Ladybug_View Analysis” component, the component assumes a human field of view in the vertical dimension that is 60 degrees from top to bottom. This comes from several studies about human peripheral vision, though the boundaries of peripheral vision are a bit fuzzy and it’s admittedly a little subjective to say where exactly it ends. In any case, the factor-of-3 difference between 20 and 60 is what gives you the roughly factor-of-3 difference between 5.555 degrees and 2 degrees.
If you disagree with this interpretation of the LEED credit, you can always change the Ladybug View component to do a 2D analysis instead of a 3D one by setting _viewTypeOrPoints to 0 - Horizontal Radial. Then the whole analysis will happen in the 2D XY plane and you should take 5.6% of the resulting vectors as the threshold for quality view.
But I think interpreting the LEED credit in terms of a solid angle is truer to the study that the credit is based on. After all, if we just gave a person a horizontal slit to look through that took up 20 degrees in the horizontal dimension, I doubt many of us would consider this a “quality view.”