> > Soft updates gives higher performance than journaling in most (not all) cases.
> 
> Can you actually demonstrate that this is the case?

IIRC these papers do.

http://www.usenix.org/publications/library/proceedings/usenix99/mckusick.html
http://www.usenix.org/publications/library/proceedings/usenix2000/general/seltzer.html