It is a matter of perspective.
Looking at your original chart at 4kHz, I would read the following estimates of the values:
Glass (single or double) E=~0.01
Plasterboard (double) E=~0.04
Plasterboard (single) E=~0.09
So that would actually indicate plasterboard is around 9 times as absorptive as glass at 4kHz.
That was the perspective behind my statements.
But yes, you are right, in the overall scheme of things, both are reflective materials!
Is there a theoretical optimum value for the absorption coefficient for a listening room?
I believe actual objective is generally to reduce reverb time.
You may be well ahead of me, but here is a link I found:
www.sciencedirect.com
If you scan down the left side you will see "(13.10)" which gives the basic function for calculating the reverb time and that "alpha" (the absorption coefficient) is on the bottom, so the reverberation time is inversely proportional to the sound absorption coefficient, and (in the same room) a coefficient of 0.01 would have a reverb time that is
9 times longer than 0.09 (single plasterboard) and 70 times longer for 0.70 (heavy curtain at 4kHz).
We can also see that the reverb time for single plasterboard (at 0.09) is around 8 times longer than curtains (at 0.70) which means (in the context of reverb time) the difference between glass and single plasterboard (9X) is equal (or slightly greater) than the difference between single plasterboard and heavy curtains (8X).
Of course my estimates of the values in the chart may not be exact, but they are close and I believe the relationship/conclusions are valid!