Filatova. This is one of my favorite sets of profiles. Simple, easy to understand, relatively stereotype free. There are a few passages that are useful for typing. I've noticed some bias in favor of delta, however (her own type is EII).
Beskova. A flattering set of profiles that is appealing on a personal level, but too riddled with all the usual misleading stereotypes such as "SEE can't math b/c Ti-PoLR". Appropriate as an introductory set only. Her profiles for her 'supervisees' were amusing when she suggested striptease as potential occupation for them.
Gulenko. I liked his descriptions primarily because he tries to derive type traits directly from his theoretical work on Model A. As such his profiles have the highest concentration of type-specific traits and the least amount of superfluous attribution. One a personal level they are a bit too dry and strenuous to read (and translate) due to being so dense in content.
Stratievskaya. Another one of my favorites. I find her profiles to be accurate for the most part. However, unlike Gulenko, she tends to attribute virtually everything to socionics sometimes to the point of absurdity. She also goes too much into describing particulars and specific situations, such that I get the impression that one needs a lot of Ni to connect it all together and derive meaning from these disparate scenarios. I think her intertype relationships articles are more elucidating of type distinctive traits than her type profiles.
Ganin. His socionics.com website is where most of people land when they start researching socionics. I remember there was a lot of debate on MBTI forums about his profiles for similar types, particularly quasi-identicals where it was suggested that he had switched the profiles around. His profiles are concise and easy to understand for beginners, but they provide a very general outline of types by which it's impossible to type accurately. After the intro stage I've never paid attention to them again.
Another author that I like that I haven't seen translations anywhere is Elena Zamanskaya. I think her observations of types are particularly insightful, but I've found only incomplete sets online:
http://www.sociotype.ru/