The idea was put forward in the 1800's that binary stars were formed in the following rather simple manner. Suppose that we have a slowly rotating cloud which is contracting to form a star.
However, what happens as the cloud shrinks?
So, the centrifugal term also strongly increases as the cloud gets smaller. We have that since v ~ 1/R ===> centrifugal term ~ 1/R**3
Even if gravity is initially dominant, because the cloud has to collapse by such huge amounts, the centrifugal term always wins unless we start with a non-rotating cloud.
This is a very appealing idea and in some form or another is probably correct! However, .... .