John, Here's another for you, though I am unsure of its accuracy! I was playing just now with the NASS Dialist's Companion and changing the date and time to find when their calculation of EoT turns to zero.
For the longitude of Greenwich (and, as it happens, 52 Lat and with other corrections turned off in case they had some effect) it gave 17.53.30 on 15th April 1999. Hmmm I would expect there to be a maximum excursion of at least a day because of the leap year problem but we shouldn't forget that the leap year correction is not just one day every four years there is a gradual build up of drift in other years as well. We are seeing that at the end of 1999 when 2000 is (unusually for a century) a leap year. Therefore the maximum drift may well be more than a day. Patrick