Need a Mail tutorial

Question

I'm trying to transfer my several email accounts from Mail in OSX10.4.11 to Mail in OSX10.6.8. I've installed my entire OSX and all software in less time than I've spent trying to get my Mail accounts across.
I need a Mail tutorial. Not one that tells you how to use it, but one that explains what all the options do and how the various mailboxes operate. For instance, I've got three versions now of each account: one at the top under MAILBOXES, one in the middle under IMPORTED, and one at the bottom named with the name I assigned it. What a mess.
Any suggestions as to a detailed tutorial that explains all this stuff?

Answer 1

Thanks Kurt for all your suggestions. I've passed a couple of Helpful Answers your way.
Since your last response a few weeks ago I've played around with Mail some more, but it became too much of a mess. There appears to be no easy way to set up a POP account in Mail unless you happen to know all the appropriate server settings. Even when you change to POP, Mail keeps calling the account an IMAP account. After 2.5 GB of unnecessary uploads and downloads (Mail was trying to synchronise, or whatever you call it, and I didn't want it to), I gave Mail the boot.
Turns out there was a one-word solution to my problem: Thunderbird. I downloaded it, installed it, opened it, chose the POP option, entered user names and passwords for five accounts, and Thunderbird consulted its database and entered the appropriate settings. Then I imported all my emails from Mail, copied them to the appropriate folders, and that was it. Worked first time.
I don't like tossing Apple software, but for me, Mail was too difficult to set up.
To anyone else who reads this thread and has difficulty with Mail, my suggestion is: try Thunderbird.
CCCloner
Re why I don't use CCC. Well, actually I do, occasionally. But as incremental backup software, it is too slow for my liking – about 1/10 of the speed of SilverKeeper (SK). CCC is slow because it makes absolutely sure that files are identical, something I don't need; whereas SK simply checks modification dates and file size. It's worked for me for almost a decade, keeping track of >50,000 of my files (backed up to 4 hard drives), and I've never lost anything.
A few years ago I contacted Mike Bombich, the author of CCC, about the speed difference. He agreed that CCC would be significantly slower than SK for the reasons listed below. According to Mike, SK uses method (b) whereas CCC uses Method (c), explaining why CCC is so much slower for my backups -- because 36,000 of my 58,000 files (that then existed) have "extended attributes" which were probably resource forks.
For the purpose of this conversation, though, the gist is that you have many files that have two data structures -- a data fork and a resource fork. Each of these data structures can change independently of each other. Resource forks, however, do not have modification dates -- the modification date of a file only refers to the data fork. As a result, backup utilities have a choice among three methods for determining whether to backup the resource fork as they scan your filesystem:
a) copy every resource fork, every time, just to be on the safe side
b) copy the resource fork of a file whenever the data fork of the file changes, or when the size of the resource fork has changed
c) compare a checksum of the resource fork on the source and target to determine if the resource fork has actually changed.
Aside from the above, the primary reason I don't clone is because an active OSX can be altered by opening it and using it, whereas a disk image of an OSX can't be changed. I prefer to make a disk image of my OSX partition, store it on my hard drives (the disk image gets backed up by SK, along with the rest of my data), and when I need to restore from it, I can be certain that the resulting OSX is pristine.
It's a bit like printing a book that was laid out in InDesign. You never send the "live" InDesign file to the printer (dangerous to do that) because the layout could change. You export to PDF from InDesign, then send the PDF to the printer. I treat OSX the same way: from a pristine OSX I "export" (clone) to a disk image, then "print" (restore) from that image.
Once again, thanks Kurt for all your help.