- I actually thought that the original deflection was pretty good.
- I suspect that the deflections are always a fair bit underestimated because, to my knowledge, the software doesn't account for joint slip at that all important bottom chord peak joint. The steep scarf cuts on the incoming pieces make for a lot of gapping and, as you can imagine, it's quite difficult to place a rectangular plate on a joint of that configuration and still get a good bite on the pieces as you'd want to.
- It's probably not a reduction in webbing that limits the deflection benefit from the 2x6 bottom chords. Rather, what's usually happening is that the deeper chord makes the truss model mathematically shallower because the chord centroid shifts. For shallow trusses, that can have a noticeable effect. In part, that's why floor trusses are 4x2 instead of 2x4.
- A great way to improve these things is to flatten out the bottom chord some. 4:12 or 5:12 maybe. No doubt that's impossible though as your architect probably has the ceiling plan telegraphing out to the exterior wall someplace and the look would be different.
- Church?