It used to be five or six votes could keep you on the front page for a while. Now you submit the same kind of article and it dies with only a vote or two, and other people coming in with the same link count as some of those votes.

Sadly, I'm left to conclude that it is not the quality of the article, but the personal relationships between the posters that lead to upvoting or downvoting. That is, the same article submitted by two different people would get highly different scores based on the track record of the people involved, not the community's feeling about the content of the article.

Speculation only, but it explains the answer to your question. We ARE seeing just as many science stories -- probably more. It's just the wrong people are submitting them. (on average)