To me, the basic vote controls (as found on Beta sites like--currently--space.se) seem the most obvious. They look very much like the buttons one might find on a remote control (upward or downward pointing play button).
StackOverflow's vote controls are very minimalistic (gray isosceles triangles). While the up and down nature is more obvious (not equilateral triangles) than at christianity.se, the fact that they are controls is less obvious than with the basic design.
scifi.se has vote controls similar to the basic vote controls, but these are still not as obviously controls as in the basic design.
electronics.se has something like a buffer gate symbol. The up or down nature of the symbols is somewhat obvious, but they probably look the most like just decoration of the above examples.
The vote controls for christianity.se are perhaps the worst of all of these examples in terms of being recognizably up and down and recognizable as controls rather than decorations. The use of a symmetric curved triangle-like shape (presumably linked to the Christian use of the triangle to represent the Trinity) makes the upward/downward nature less obvious. The highly ornamental nature of the images also makes them less obviously control buttons (rather than decoration).
However, if one can guess that the number is, in fact, a score from voting, it is not that difficult to discern that the image above the score is for up voting and the image below the score is for down voting. (Even if one mistook the number for an article number or rank in a series of articles, one would probably guess that the buttons are previous and next controls from context.)
If one guessed that the number is a score, one might think that ordinary users do not have the ability to vote, so the lack of clarity has a small disadvantage.
Recognizing the voting controls is probably a much less common problem than recognizing that an asker can accept an answer.
With respect to few people reading the about page: while this is true (as noted in comments, only 179 Informed badges), the fact that 1,354 Supporter badges (first up vote) have been earned might indicate that the problem is not that severe.
Here are some numbers (with sites ordered by my rough ranking of most decorative/least obvious voting controls):
SE site | Informed badges | Supporter badges | Users
| (% users) | (% users) |
| | [* Informed] |
christianity | 179 (5.21) | 1,354 (39.4) [ 7.6] | 3,435
electronics | 398 (2.10) | 6,080 (32.2) [15.3] | 18,910
stackoverflow | 52,587 (2.47) | 419,292 (19.7) [ 8.0] | 2,131,780
scifi | 326 (2.97) | 4,503 (40.1) [13.8] | 10,988
space (Beta) | 79 (9.55) | 402 (48.6) [ 5.1] | 827
There is probably a significant bias against earning the Informed badge on a secondary site (if one already has significant experience with another SE site, the about page is less interesting). Users in a early public Beta stage are probably more likely to up vote (so the high percentage of Supporter badges in space.se might not have much to say about the obviousness of the interface). Being the most mature and popular, StackOverflow may get a larger fraction of low-participation users (possibly somewhat explaining its low Supporter count).
I suspect that site age has more to do with the percentage of users that have up voted a post than the obviousness of the voting interface. (Ordering by age, I think scifi.se is the only one of these sites with an anomalous percentage of Supporter badges. That might be somewhat explained by the more subjective content of the site and perhaps the more enthusiast traits of its users.)