In context, the move to SHA256 from SHA-1 (128) is probably intended to decrease the likelihood of a hash collision, readers may find the following link of interest on approved crypto strengths and crypto-material-at-key-strength retention periods.
http://www.cse-cst.gc.ca/documents/publications/itsa-asti/itsa11d-eng.pdf
Accordingly, SHA-1, as a crypto component, is disallowed as of 2008/12 by the above Canadian govt standards.
It's interesting to see what algorithms and key-strengths are deprecating in 2010.
/S