I have checked many of my Garmin 496 'captured' waypoints on Google Earth and find them to be easily within 3 metres accuracy and probably better than that. My 496 and panel mounted Garmin 530 are both EGNOS enabled and so, as you are no doubt aware being an IT expert, this is verified by ground station down to an accuracy of 1 to 2 metres (see
www.esa.int/esaNA/GGG63950NDC_egnos_0.html ).
For the record, I use GPS as a primary form of navigation all the time (with EGNOS enabled) except for instrument approaches in IMC. VORs (when they are working) are only used as secondary cross checks when enroute.
I have also emailed Google, out of interest, and asked them to comment on the Lat/Long accuracy of their Earth product. However, quite apart from my own verification, it is obvious from the user guide that accurate location and distance calculations are a fundamental basis on which the product is built (see
http://earth.google.com/userguide/v4/ug_measuring.html ). It is often used in conjunction with a GPS receivor as a moving map so it really is not the 'toy' you seem to be suggesting. I will let you know their response.
You stick to your map and clock if you want to (and the 30% of the year that this technique allows you to fly) but please be aware that the merry band of GPS whingers is thankfully in decline. (and if the fifty or so of you left feel inclined to have a rant then please just start your own thread so that I can avoid it )
