Some thoughts on statistical projections of the WTBTS membership count:
1) Extrapolation (projection beyond known statistics) is hazardous for any measurement that does not derive from a simple model. An accurate model of the WTBTS membership count would have to include demographics for each of the countries, and this would have to combine population age distribution data, educational attainment data, economic data, current WTBTS coverage, and (of course!) Internet access availability data.
2) Some countries are easier to model than are others. Countries in westerns and northern Europe are all relatively stable, wealthy, highly educated, and have (for the most part) homogeneous populations with a fairly constant age distribution. The trends in these countries are obvious (multiple consecutive years of baptism declines), and so are relatively easy to predict.
3) Some countries are much harder to model. These are those like the former republics of the Soviet Union where the WTBTS influence is new to many of the inhabitants and the cultural acceptance is difficult to quantify. An example of this is Japan of several years ago when the WTBTS numbers were growing significantly. How many people could have correctly predicted the very steep decline there of the past few years?
4) Modeling is more difficult in countries were the WTBTS is faced with active competition; these include many of the nations in South America. Gains made by the WTBTS may be due more to shortcomings of its competition than to any special efforts by Brooklyn.
5) A very big lure of the WTBTS is the promise of a paradise Earth (Real Soon Now) and the appeal of this is undeniable among the poor is less developed and economically stratified countries. Modeling this component is hard because it subsumes a general economic model.
6) The game theoretic strategy of the WTBTS itself is not well understood and is subject to change. They have many types of adjustments they can make, but two stand out: more money/power vs. more converts, and more short term gains vs. more long term gains. The WTBTS may employ a mixed strategy that is dependent on the specifics of each country.
7) There is the doctrinal wildcard: it is possible that the WTBTS just might try another time prophecy for the arrival of the Big A. The 1975 debacle is unknown to many in the developing countries and is apparently accepted (or ignored) by many of the members in the more prosperous nations. Perhaps there are other doctrinal changes they could try.
8) The geopolitical unknowns: what if, by some miracle, communist China became openly accessible to the WTBTS? The world total membership count could double in a year and maybe continue doubling for several years.
Anyway, my prediction is that the global membership count will continue to increase for the next four years, peak in 2006, and start a slow but sure decline through 2011. After those years of decline, we can expect the WTBTS to mutate as it has in the past, although the changes involved are anyone's guess. Perhaps as a last attempt to stop the inevitable decline, they just might try to go mainstream (no goofy time prophecies, no shunning, no "God's Only True Channel", real charity, real ecumenicalism, etc.). It will be a desperate tactic, for they know that once they try it, there's no turning back.