One problem I see with the approach you suggested is that work is already visible to many but may or may not be accepted in the end. Publications are often rejected not due to a lack of innovation but because of other aspects (evaluation, comparison to other work, ...). A new idea appearing in this paper will then be available to the world while the author is not (yet) credited for his work. Someone else might just pick up his idea, fix the problems, and publish it himself.