[viggio24]

You shouldn't consider this as a TM failure but just as an HDD failure. You may have some very o,d Snow Leopard backed-up data (e.g. parts of the OS which are never updated) or several months old documents which have been saved once by TM and never touched again. So TM cannit be blamed for such old filesystem nodes which have been damaged by the HDD failure and not TM one. What you tried to do is to completely wipe your HDD and do a complete restore by TM. This operation should be better accomplished using a fresh carbon-copied HDD image, that ensures all data are new and safe. What you can ask to Apple engineers eventually is to add a "cold restore" operation, that is asking to do some sort of simulated restore in order to check needed data integrity only, thus preventing users to make risky operations such as the one you tried to do.

[Vitaly]

How does TM know if it can reuse file from the previous backup or should create a new version. If it's just timestamp then it's not such a good idea, it should use hash of the content, and if it has and stores content hash it can and should verify backups once in a while to e sure against bad sectors etc. Doesnt have to be on each backup, and doesn't have to be all files at once, but it can, say, check backed up file content once a month/week or so.

[hiraasht]

Thanks for the comment viggio! Never considered it as a failure of Time Machine but I sure think it is something that can be prevented by Time Machine by just running a integrity check on the disk drive (even as optional feature) prior to doing a backup. After all its a backup that you rely on :)