OK well as expected this falls into the category of YYGT (Yes You Get That)
Seriously though a reading of 5 and a max of 3 and you get greater than 100%. This happens to most folks whether or not they have the solar sensor shaded or not. The design of a Davis solar sensor is one which measures global solar not direct, that is it reads the whole sky including any reflections etc. This makes life difficult when you are estimating the sunshine as a % from actual w/mē values. Based on the values on those screen shots it seems likely that you need to stop WD setting sunny for a minimum of 30 minutes, my calculator suggests a reading of about 36 as expected max after around 30 minutes and you would need a actual reading of around 31 to show sunny with an 85% threshold.
As a slight aside to this I am puzzled by your lat/long values as they show up on a map as being some distance away from Daventry although not enough to make a significant difference to the expected max calculations. Daventry shows up on google maps as being about 52.2575N (52° 15' N) and 1.1649W (1° 9' W) although I realise you are probably not in the centre of Daventry your values put you north of the M6 near Churchover, which is correct?
Edit. I just checked back and you have a sunny threshold of 70 which is way too low, it needs to be 80 at least and preferably 85 with the current way WD does this calculation. I have 85 sunny the 65 45 25 and lastly 15 as my values from sunny down.