I know that, in general, the higher-ranked manga tend to be shounen, such as One Piece, Bleach, and Naruto (based generally, I think, on readership). Is there anything showing general demographic popularity, such as the overall popularity of shounen vs. shoujo vs. seinen vs. josei, based on something like readership? If so, is shounen still the most popular? How do the other ones rank in relation to each other?
This is a bit hard to measure, but shounen is clearly the top demographic group. Popularity is usually measured by magazine circulation. The difficulty in this is that there is a lot of overlap of readers, and also that different demographic groups read widely different amounts of manga. For instance, a young reader in the target audience of shounen magazines is likely to subscribe to only one or two magazines, while older readers might subscribe to many. Thus, total circulation numbers for the older demographic groups tend to be higher relative to the number of readers.
╔═════════════════╦══════════════╦═════════════╗ ║ Demographic ║ Top Magazine ║ Total ║ ║ ║ Circulation ║ Circulation ║ ╠═════════════════╬══════════════╬═════════════╣ ║ shounen (boys) ║ 2,876,459 ║ 8,344,534 ║ ║ seinen (men) ║ 807,871 ║ 7,624,811 ║ ║ shoujo (girls) ║ 745,455 ║ 2,803,230 ║ ║ josei (women) ║ 162,917 ║ 2,364,666 ║ ╚═════════════════╩══════════════╩═════════════╝
It's pretty clear based on this that shounen is the most popular, followed by seinen. Shoujo is less popular than seinen, but the numbers are similar for the top magazines, suggesting that the fanbases are somewhat similar in size. Josei is less popular than the others. However, in terms of total sales, seinen is almost the same level as shounen, and josei is almost the same level as shoujo, suggesting about equal market shares.
Admittedly, the demographic listed for the magazines is not perfect. However, I think the above is a pretty fair representation of roughly what the sizes are of each of the four main demographic groups. Also, this data is from 2010, so the current data could have changed a bit, but not enough to change the conclusions. I used 2010 data because it was easily available in English and formatted in a way that made calculating totals easy, but more recent data is available through JMPA's website if you follow the link above.