Do Graphic Vinyl Banners Affect Sound Transmission?

I've done a bunch of outdoor sound systems over the years, and often times the event organizers (the always excellent Good Sense & Co in this case) want the speaker stacks covered with some client-related graphics.  Last spring, we had Elton John at the Tribeca Drive In:

 

Behind those graphic panels left and right are Meyer MILO arrays:

On these events I usually have just barely enough time to get things in the PA roughed in before the rain arrives or something else happens, but listening, it never seemed that this vinyl affected the sound all that much. If anything, I sometimes put a slight high shelving filter boost on the array to compensate. 

But after that Elton show, they were tossing out the vinyl, so I cut a piece of it and took it back to CityTech, where I ran a little SMAART on it a few months ago (I'm just getting time to write it up now).  Working with a couple student volunteers, we hung the vinyl (which apparently came from Color Reflections is just called "85%", meaning 15% open) in front of a Meyer CQ-1 and played pink noise:

(The AE-10 subwoofer at the bottom wasn't wired up)

After straightening out yet another Dell grounding problem, we measured it in SMAART spectrum mode with a five second average both with and without the vinyl, and not changing anything else. This screen capture shows both traces; the red (in front) is the measurement with the vinyl in place; the green is with the vinyl removed (Click for larger images):

I took both plots and exported them into Excel, and made (with some help from Calvert and Jamie at Rational Acoustics) this plot:

Keep in mind that this was based on two spectrum measurements taken at different times, so it can only give a general sort of average response, not a clean transfer function of the vinyl material (I didn't have time to compare two microphones and then apply that, etc etc).  Also, I cut this chart off at 100Hz since there wasn't a whole lot coming out of the CQ down at those frequencies (the measurements went down to about 1Hz) and there was some low frequency variable background noise (subway) and not a lot of data points, making that area of the chart kind of meaningless.

But what I feel I can get from that graph is a sort of general trend of a very slight 1-3dB roll off starting about 6kHz, and extending way above what I can hear.  So, I think my strategy of a gentle, slight shelving boost made some sense, if only to brighten things up a bit, even though, from my experience, the wind, humidity, and other factors outdoors have a far greater impact on the quality of the sound than these banners.