I have issues with the 'Wood window with storm' option.  I evaluated a 1938 home with original wood single-pane sash windows.  It has full frame wood insert windows that fit on the outside of the house for winter.  I consider this, for the purpose of the Tech Procedures, to be a "Wood window with storm".  However the stats (for a 23in x 36in window) look like this:

ER-55;  R-Value: 4.68;  SHGC: 0.4521

A modern window, Double/double with 1 coat; Low-E .20(hard); 13mm Argon; Insulating spacer; Hinged, has stats like this:

ER-20;  R-Value: 2.76; SHGC:  0.4666

This is causing upgrades to create negative performance gains. The 'metal window with storm' option has an ER of 63, but at least the R-Value is 1.24

For now I am going to model this as 'Wood;double;clear;insulating;Slider with Sash'  This produces:

ER-9;  R-value: 2.06; SHGC: 0.4560

Any opinions on this conundrum?  Am I missing something?  Is there a better way to show clients how much window upgrades help, in a more accurate fashion?

Well, it certainly seems the "Wood widow with storm" on your computer comes up with much different specs than mine.  I like your solution and it is essentially the same as one the suggested by one of my colleagues.  However, the H2K user guide says we must use the pre-defined codes.  Perhaps I have a corrupted code file?