Actually GPS has shown a number of smaller islands to have been historically mis-plotted. Most of these are now fairly well known and charts have been updated.
I'd point out via Bob's excellent example that, when it comes to cartography, GPS, and visual/radar navigation the visual/radar is to be assumed the most trustworthy. The value of the coastal chart is not in it's precise correlation of terrestial masses with an accurate latitude and longitude but in the accurate depiction of their relative locations. In simpler terms, once landfall is made latitude and longitude cease to be of great navigational significance. The island is exactly where you see it as being, either visually or on radar, regardless of whatever the discrepancies may be in assigned latitude and longitude. Perhaps the best correlation may be made with the fathometer, or depth sounder, versus the charted depth. Which are you to believe? (After all, cartographers and surveyors have off days also!)
Ignoring the effects of radar super refraction, which manifest themselves only at very long ranges, usually within a temperature inversion, the radar is to be trusted above the GPS. I'd have been inclined to think, were I in a similar circumstance, that the GPS was "broken" not the radar.
If I'm reading your display correctly, and there is some doubt in my mind as to that, the error is on the order of minutes not seconds, ie.. miles off. Are there more islands off screen that I'm not seeing and what is the nature of the topography of the islands. I'm having trouble resolving the two images I'm seeing as being the same-but then, I guess that's the whole point, eh? (g)
“Scientists are people who build the Brooklyn Bridge and then buy it.”
Wm. F. Buckley, Jr.