A great feature would be for Google to retain the checksum of the page at the time it was indexed and pass that along to the browser (eg: Chrome). If the check

A great feature would be for Google to retain the checksum of the page at the time it was indexed and pass that along to the browser (eg: Chrome). If the checksums don't match, the browser can do a background call for Google to re-fetch the page and recalculate the checksum. If the checksums still don't match display a warning message to the user. There would likely be some work required to weed out false positives from things like dynamic content, so perhaps only non-text parts should be checksummed... but if one of the biggest issues is malware sites serving up an innocuous page to Google crawlers instead of the malware-filled page that is served to users, something like that would be a good initial sanity check.