This is a fun little exercise, but I have very little faith in it's accuracy. They left out some major belief systems that might've been interesting, too.
1. Secular Humanism (100%)
2. Unitarian Universalism (92%)
3. Liberal Quakers (79%)
4. Nontheist (70%)
5. Neo-Pagan (70%)
6. Theravada Buddhism (69%)
7. Mainline to Liberal Christian Protestants (61%)
8. Taoism (57%)
9. New Age (50%)
10. Orthodox Quaker (47%)
11. Mahayana Buddhism (47%)
12. Reform Judaism (44%)
13. Jainism (38%)
14. Baha'i Faith (37%)
15. Sikhism (37%)
16. Scientology (31%)
17. Church of Jesus Christ of Latter-Day Saints (Mormons) (27%)
18. New Thought (27%)
19. Seventh Day Adventist (25%)
20. Hinduism (23%)
21. Christian Science (Church of Christ, Scientist) (22%)
22. Mainline to Conservative Christian/Protestant (20%)
23. Eastern Orthodox (20%)
24. Islam (20%)
25. Orthodox Judaism (20%)
26. Roman Catholic (20%)
27. Jehovah's Witness (13%)
It might be nice to see a longer test that starts by narrowing things down, then gets more detailed to draw more effective distinctions among the top choices. I think that would be a more useful test, and it would probably produce more intuitive results.
I'm just laughing at this. Seriously, Islam is number 3??
This is exactly what I'm talking about. Islam is an Abrahamic religion that shares most of its basic ideas with Judaism and Christianity, so in a test like this, Christians will score higher for Islam than for certain sects and denominations of Christianity. After a few questions, the test could determine that Steph's beliefs are in line with those of the major Western religions, then proceed to more specific questions like the importance of Jesus, etc.
I realize the test was deliberately avoiding that to some extent, trying to give people a feel for which religions most reflect their spiritual philosophies and moral beliefs, but it seemed like the test couldn't make up its mind about exactly what it wanted to be. I think you need to commit to one side or the other: make a test that's specific and primarily produces the expected results (or sheds light on appealing new possibilities), or create one that's exclusively focused on moral ideas and very basic questions (How many gods? Zero/Don't know, One, More than one) to tell people what their religion "should" be.