Indeed.

You cannot compare a surface wind forecast in a TAF with the wind forecast at altitude. Obviously at low altitudes they will be pointing in roughly the same direction, but they will often be much stronger than at the surface.

TAF forecasts contains winds at ground level doesn't it? The plog will contain the wind for the level you've planned to fly at.

I also had some issues today with the wind.

It would be a mistake to call one forecast more accurate than another; they are both just forecasts. One day the GFS model output might more closely reflect actual conditions, on another day, it might be the local Met Office model output. I would expect on average the local forecast would more closely match, but the differences on average should not be significant for VFR operations.

I was comparing the winds in SD against F214 this weekend and there was a considerable (20kts +) difference @ 3000'

Has any work been done to verify if GFS is more or less accurate than Met Ofice forcasts for the current day?

Winds aloft forecasts are sourced from the Global Forecast System.

I am currently doing my IMC rating, so a lot of time spent working out drift corrected headings etc. Over the last few weeks, I have found the Metoffice forecast wind charts (Form F214) to be significantly more accurate than the winds aloft shown in Skydemon (where I would normally set a precise take-off time). What is the souce for winds aloft in Skydemon?