Inactive Players Missing

My post on the Growth of the FIDE Rating System identified a number of anomalies in the data. These are years where the count of players declines from Year(N) to Year(N+1), then jumps from Year(N+1) to Year(N+2). For example:-

2001 - 36.977 (my original table had 36.976; one anomaly at a time)
2002 - 29.283
2003 - 45.015

Since FIDE IDs are available in the data starting in 1999, it's easy to match players from one list to the next without the headache of trying to match names. After a short investigation, I determined that 11.314 FIDE IDs were present in the 2001 & 2003 rating lists, but not in 2002 (I'm one of them: FIDE ID 2003740). Looking at the 2001 data, 11.008 players had no games in the preceding period and 7.055 were marked inactive ('i'). It appears that FIDE dropped inactive players from the list at some time in 2001, then reinstated them in 2002.