I remember well meeting some recorders in Wicken Wood on the borders of Bucks/Norhants who were logging Wood White distribution using GPS.
During the conversation, I queried whether they were aware of some trick (which I had learnt about the hard way when I got it wrong) to ensure that the right database was used to enable the GPS to read OS grid references. I cannot remember the exact details of the problem, but if the wrong database is use, the grid ref is about 300 metres in error. So, having agreed from a printed OS map, exactly where we were in the car park, they cross-checked their GPS. Lo and behold, there was an error of some 300 metres. So that raised the question: which county had their sighting been in? It all seemed a bit academic as Wood whites clearly occur both side of the county boundary, but an interesting GPS problem nonetheless.
I much prefer Street Map:
http://www.streetmap.co.uk/ (enables arrow to be positioned precisely) to Get a Map, but that is a personal preference.
Jack