I don't have any experience, so I can only supply a data point.
The attached picture shows the location of the WH32 in the shield and hanging on a string an old-ish temperature sensor for a radio-controlled Oregon Scientific alarm clock. In addition I put a WH31 on the railing (?: English) to cross check whether the Oregon sensor might be "in a sea of cold". The balcony is facing north-west, the picture is taken at about 2:30 pm local time, and the weather underground station name is IFSSEN8. The sky was more or less clear all day. Before I had the WH32 and the shield I had a WH31 hanging next to the Oregon sensor; they were typically less than 1 ° C apart.
Today the Oregon sensor never went above 1.5 ° C; most of the time it was around 1 ° C. As you can see on weather underground, the WH32 inside the shield reached 8 ° C. The WH31 was not out there all of the time, but after a while, it tended to agree with the Oregon sensor. So, unless the WH31 also suffers from a sea of cold effect, the WH32 inside the shield and in November (I'm in the Northern hemisphere) is still picking up some solar radiation.
I'd ask other, experienced people to chime in and say whether - based on the data I gave - the shield is doing a reasonable job or not.
Personally, I have to say that I was surprised about the temperature difference between the WH32 in the shield but with November sun, and the Oregon sensor without sun. I am considering moving the sensor to the right side of the balcony (a similar beam is located there; actually it's the "middle beam" of the balcony as my neighbors' balcony is right next there, separated by a wooden panel). I guess the sensor should pick up less sunlight there, but it may be more affected by radiation from the building. If anyone has any thoughts on that, please share.
HTH