You raise some interesting points--I had to hold off responding until I had time to tabulate a little more data, since arguing "that doesn't seem right" when you're talking about population statistics is silly at best. Below is the table I came up with from jcwren's stats pages.

In case the headings aren't clear, the cumulative total is the number of monks at that level or higher, which is (generally) equivalent to the total number of monks who have ever been at that level; the fraction not yet advanced is the portion of that total who still are at that level.

Ignoring the last few rows for now, it seems your point about the high fraction of pontiffs moving on to sainthood is a valid one. I wouldn't put it as dramtically as you do above (though those numbers are certainly correct): roughly 65% of bishops have gone on to become pontiffs, while 74% of pontiffs have gone on to be saints, which seems a slightly more apples-to-apples comparison. In general, this seems to be a trend from level six onward: each level from six through nine has a lower retention rate. This is consistent with my theory, but the large drop at level 9 is still out of line with the trend, which lends some support to yours.

Of course, this analysis leaves something to be desired, since it only looks at the levels as black boxes, rather than what they are (points on a number line). I don't think it alters the conclusion significantly, but for the record, the percentage increase in points required to move from each level to the next is (starting with 2 -> 3, for obvious reasons) 150%, 100%, 100%, 150%, 100%, 60%, 44%, 30%. So in the higher levels, each transition is, in fact (relatively) easier than the previous one. Though that last step is a doozy. ;-)

In any case, thank you for the reply, it's always good to be forced to check your numbers. :-)