Copying the entire drive is impractical. Not only does it take way too long, it makes restoring lost work much more difficult than it would be if you selectively backed up your own data files. IME most folks backup methods don't really consider the process for restoring lost data.
BTW for Outlook, there's a file called outlook.pst hidden in your "documents and settings" folder tree that contains all your local email.
I'm not sure I understand.
Casper does take as much as a few hours to make the first clone of a very large hard drive, but subsequent copies only take a few minutes because only the changes are copied.
As far as restoring, the recovery process in the case of hard drive failure consists of swapping the drive and booting up. There's no compression, hence no extraction. It's a bootable clone.
In the event that you lose or mess up a single file or folder, the recovery process consists of navigating to it's copy on the backup drive and copying it to the primary drive.
This is not to say there are no disadvantages. As Dr. Bruce pointed out, the possibility of copying over viruses or system flaws is there. There's also the possibility of losing both drives due to a power event, a bad HD controller, fire, flood, theft, etc.
In addition, the Casper clone is of the drive as most recently copied. There's no way to go back to a date before that. (Well, okay, I guess you could if you wanted to use multiple drives and alternate them like we used to do with tapes; but that's a bit more than most people are likely to do.)
So no, Casper's not the end-all. You still should have data backups (preferably off-site via the Internet or on removable media that's stored off-site). Also, Casper won't help you if the backup drive gets hosed along with the main drive. That did happen to me once a few years ago; although even then, even though the system was corrupted on the backup drive, I was able to use it to recover the user data, downloaded software, and so forth.
In essence, what I'm saying is that Casper is a very quick and easy way to make up-to-date, differential clones of a hard drive that make recovery a breeze in 99 percent of events that cause the main hard drive or the system to get hosed. The key is to clone to a drive that can be physically installed as a direct replacement.
Yes, you can also use the Casper boot disk to boot to an external drive and copy it back to the system drive, but what's the point? Why not just use a physically compatible drive to begin with and lose the extra step?
But hey, everyone has their own preferences for backup, and it truly does my heart good to see threads like this one where the necessity of backup is taken as a given, and only the relative virtues of different backup methods are debated. FWIW, this is my typical minimum backup strategy for SOHO clients:
1. A second, physically compatible drive is installed on every workstation (internally or externally), to which the system is automatically cloned using Casper on a schedule appropriate to the client. I usually schedule it for the late afternoon (right after closing time), so if any problems show up on the system during the day (viruses, etc.), I can either take care of them or tell the client to shut down the machine before leaving so the clone will not happen that day.
2. Offsite backup of critical data is made daily or more often to
FilesAnywhere (or to a protected directory on their own Web server or another remote computer that they own, in some cases). Once the initial backup is made, subsequent backups are either incremental or differential, depending on their needs, and take only a few minutes over a high-speed connection.
Handy Backup is a good, cheap, simple tool for this. I like the fact that I can set it up to notify me by email when a backup has occurred. Every morning I check a special email address to make sure all of my clients' backups have occurred. If not, I investigate.
3. If the client has a server, it's RAIDed with a hot spare, and critical data is backed up remotely as above. I also like to make incremental backups of the system array in case it's necessary to go back a few days, using whatever method makes most sense given a particular setup.
It's not a perfect strategy, but it's not bad, either. There are other things I'd like to do (periodic archived images of workstations, etc.), but the above is a minimum strategy for small businesses and is designed to balance economy and redundancy, while still providing ease of implementation and recovery in the great majority of cases.
Again, though, I'm happy to see all the recent threads about backup. It does any IT guy's heart good to see backup happening, by whatever decent method.
Rich
which is more than 99 percent of people do. 
