I've just had a check of this, and that fillet should not have been present in the SkyDemon chart. Unlike some others, which are correct, that one is our mistake. We will get this corrected for the next chart update. Many thanks for your help.

Is there a view as to which is the 'official' version? Ie. in the Southend example above, if I fly along the edge of the airspace shown on the CAA chart (e.g. northbound from Folkestone), I will be fine, whereas SkyDemon will tell me I am infringing the little bit of the airway poking out beyond the line shown on the CAA chart.

(I appreciate this Forum is probably not the best place to give me an 'official' answer on this!)

There are a number of such fillets which technically exist but the NATS chart has chosen not to show. It is an unfortunate consequence of the way the UK defines most of its Class A airspace (implicitly). SkyDemon goes by the book, and the book states that those fillets exist and are Class A airspace. In the case where the AIP states non-standard base level changes we always attempt to manually intervene and alter the shape, though.

The issue with an airway meeting the FIR boundary is slightly different. We do not currently have a practical way of making our Class A do what one paragraph in the AIP states that all airways should do, which is to naturally blend themselves in to the edge of the FIR instead of having a flat edge as they do.

There is quite a bit of that near DTY VOR and Cranfield.  Change the level above which Skydemon hides airspace and the differences become really apparent.

Apparently its because the CAA chart shows quite a lot a non-standard step downs that aren't described in the textual descriptions of the airways. 

It looks like SkyDemon draws a box around airways, which causes some differences compared to the CAA charts once you get to the end of the airways.
Compare L10 here on SkyDemon vs. paper chart:
  

And Q41 near the Channel Islands:
 


Any thoughts?
Thanks
Phil