I get Richard's point in as much as simply comparing two years in themselves does not prove anything. For example, saying that (hypothetically) in 1965 there were 1m witnesses and they grew by 100k in a year and in 2015 there were 8m witnesses and they grew by 265k does not prove very much - the second number is a smaller percentage increase but there is no real context to provide any indication of what the difference really means.
When you look at it over a period of years, however, the changing ratio starts to tell a story. If you add to that statistic further data, for example the number of hours spent to get a ratio for hours of preaching to gain 1 conversion, or a geographical breakdown that shows the increase is virtually static in developed nations then you can start to formulate some conclusions.
Cherry picking a couple of years for comparison does prove very little and I would agree with Richard that one it's own even a reducing ratio of year on year increase does not really prove anything of significance. Regardless, when you look at numerous metrics taken together then I agree with all the posters that say the percentages give you the real story. The evidence presented by even just a few related lines of analysis is damning and is 100% the reason why the WTS concentrates on the headline numbers, the long figures that present an impressive picture until you look at the trends.