I know, I’ve seen them both :-)
As you correctly mentioned, the top100 rarely changes and thus has limited practical usage. I’d like to see something more dynamic and relevant for us plebeians, i.e. being able to sort all users by karma, comments, submissions, creation date.
Among other information from Whaly: there are roughly 150k active HN members per year. The 10k list gets you by far the overwhelming lion's share of activity (it's already highly concentrated in the top-100 list).
I don't know that there's a comprehensive member list available overall, and the lack of some sequential userIDNumber identifier means that the space isn't readily searchable. I suspect Whaly's approach of snarfing all activity via the API is probably the most comprehensive. A new search tool might be able to do that (and tie in other metrics such as posts and karma) as well.
There's the additional issue that overall karma is not a particularly useful measure. It's possible to achieve a high karma simply through excessive activity (my own account is a case in point). There are domain experts such as Alan Kay (@alankay) who have decidedly pedestrian karma (4,566), but that over 1 story (1,400 votes) and 110 additional comments, or about 29 votes/comment. That's well above my own average, which is about 1-2 votes per comment (2,087 stories, 28,107 comments).
Karma/posts / karma/comments might be slightly more informative, but can also be skewed by submitting a highly-popular story. The larger problem is that truth, expertise, and credibility aren't popularity contests.
A further problem with expertise is that it is highly domain-specific. The fallacy / error of ascribing general expertise to one who's shown mastery in one specific area is a grave one. Particularly when that one specific area is earning money / winning a lottery.
Oh: and based on my front-page scraping, I coded up some command-line tools to return activity / results for domains and members so I could do quick checks on both.
Note that what makes the front-page is an interesting function of both what the member posts and what the membership as a whole responds to. Someone might submit a large number of unappealing topics, but get traction on a much smaller subset. FWIW that seems to be my own experience. My tools will show the posts that got traction, but not the much larger low-vote long tail. Or flagged/killed posts, FWIW.
