@TaylorRobert 's point is correct. The U-factor and SHGC must represent the entire fenestration system (both opaque and glazing portions of the window). In addition to using values from Window 6, you can use the NFRC ratings on actual fenestration products.

It is worth noting however, the simple glazing system takes a fairly simplistic approach in that it does not explicitly simulate multiple panes of glass (e.g., reflections between the panes), any special coatings on the interior or exterior of the window (e.g., low emissivity coatings), or any opaque components of the window (e.g., frames and dividers). All of these effects are lumped into a single, monolithic layer with the same U-factor and SHGC. The simple glazing system does do a fairly good job of simulating the angular dependence of solar heat gain.

This only guarantees similar performance at the NRFC rating conditions, though it is likely to be fairly representative under a wider range as well.

Also, if you're doing more detailed daylighting simulations using Radiance, and the precision of the angular dependence is important, you may need to consider more complexity of the glass itself. Refer to this post: https://unmethours.com/question/1037/...