In the latest issue of the geo: magazine I stumbled on an article titled "Are WGS84 and ETRS89 really identical?". Though the article is basically an advert for an application to convert between the two coordinate systems, the interesting fact which I wasn't aware of before is that additionally the the coordinate system WGS84 used by GPS there's the ETRS89 used in Europe simply to avoid the change of coordinates due to continental shift. With an annual movement of the Eurasian Plate of 2.5 cm the difference between the two coordinate frames will grow by one meter every 40 years. As GPS has an accuracy of about 10 meters, the difference between the two systems is not that relevant for most applications, though there are of course some few where a few centimeter make a difference.
While I haven't found values for the plate movements relative to WGS84 for any location in Thailand, since (southern) Thailand is on the Sunda Plate together with Singapore, the movement of an observatory there can give an idea for the speed further north. There it has a movement of 6 mm per year in latitude and 28 mm in longitude, a similar speed value as for Europe, so also there it's about one meter every 40 years.
Now to relate this with the administrative subdivisions, in 2008 I found a marker within the administrative center of Surat Thani, located between province hall and provincial court. This marker shown in the photo has inscribed the location of itself, both with longitude and latitude in degrees as well as the coordinates in the UTM system. Due to the reflections its not easy to read in the photo, but the values give a latitude of 9° 7' 58.32484'' North and a longitude of 99° 19' 53.90800'' East in WGS84. The UTM values are a Northing of 1,009,256.493 and an Easting of 536,772.956 in the Indian 1975 frame. As the UTM coordinates are measured in meters one can directly see the accuracy of the coordinate, which is 1 Millimeter! 0.00001 arcseconds at the equator correspond to 0.3 Millimeter, so even more accuracy. Given the above words of plate movement, as well as the inherit inaccuracy of GPS both values have a ridiculous accuracy, anything below one meter I cannot believe. I really wonder if any geographer was involved in the creation of this marker and how they came up with the last digits of these numbers, or if they were pressed by some official who though more accuracy is better, and any reason they last digits are simply nonsense were overheard.
When I was looking for the location of PAO office in Trang, I noticed that in panoramio someone had uploaded photos of a very similar stone located right in front of the PAO office, on the backside of the province hall. On the detail view one can see the same ridiculous accuracy of the coordinates. Though I had visited a few more province hall, I haven't noticed such a stone at any other than Surat Thani; would be interesting to know which other province halls have one, or even get some more background on how these markers were created.