(That said, innovations that allow people to write more papers or get cited more have been taken up very rapidly.)
This has had the pernicious effect of increasing politics and popularity costs in, at least, the physical sciences. I know several research scientists with several hundred publications to their names which these 'authors' have not even read.
If you can solve the problem of requiring scientists to amass a large publication set by any possible means then you'll definitely change scientific publishing. No-one really wants their papers to go unread due to an access fee. That's why www.arXiv.org was invented. However, the only solution I can think of is to separate science into management and technical (this already happens to a certain extent in High Energy Physics) and allow the heads of group more freedom in choosing who to hire, and how.
This requires a bit more trust in the scientists and reduces jobs in the administration, so I fear that it will never happen, alas! Furthermore, you'd need to make both streams formally equal in order to remove the suspicion that scientists become managers because they don't have sufficient technical skills.
Both these hurdles can only be cleared with social solutions though; technical solutions can only go so far.
Now I am not completely familiar with academia, so I was wondering if you could provide some context. Is is blatant plagiarism of other scientist's work? And if so, is this something that most scientists can reasonably expect to get away with if they tried. Or is it something where the scientist has some link to the work, such as heading a lab or working with the main author, but just has not done much work on the paper?
I was fortunate enough to work in an area of research for which the arXiv was originally established, and found that printed journals became increasingly irrelevant to me. Certainly, I would use them to find the dwindling number of papers not covered by the arXiv for reasons of age or subject matter, but for the day-to-day business of keeping up with the latest developments, the latency between submission to and publication in peer reviewed printed journals is far too great to be useful.
These preprints are not formally peer reviewed, but this mostly doesn't matter since they are always subject to an informal review by their readers -- if you are reasonably well-versed in a subject you can make up your own mind about whether or not a paper is any good, particularly after discussions with colleagues.