I tried out JungleDisk as well as SpiderOak as I was looking for client-side encrypted cloudspace.
I was disappointed in both of them.
Spider-Oak was slow and I could never tell what the software was actually doing. JungleDisk was more transparent. Yet both have severe issues with special characters in filenames. I'm from Germany and we have umlauts (ä,ö,ü) and other weird characters (ß) which are part of our language. When cross-syncing between a PC and a Mac I ended up with duplicates for all files with special characters, half of them not accessible anymore. It was a big mess. I thought these special character issues should be gone, but they are not.
I went back to DropBox and now use Securstick (http://www.withopf.com/tools/securstick/) for client-side encryption as I wanted individual file/filename encryption instead of a fragile encrypted container (although plain Truecrypt works fine with Dropbox).

Duplicati (http://code.google.com/p/duplicati/) with plain S3 is also a contender for client-side encrypted cloudspace, but it's not out for Mac yet.

AFAIK TC Explorer needs admin rights as well.

Just a week ago the excellent c't magazine introduced a tool for true usb-stick encryption with open source tools, working in Windows/Linux/OSX.
This method does not need admin rights and works cross-platform.
Check the (google translated) website here: http://translate.google.com/translate?u=http%3A%2F%2Fwww.withopf.com%2Ftools%2Fsecurstick%2F&langpair=de|en
Or Original in German here: http://www.withopf.com/tools/securstick/

already a couple of months ago I discovered ndff, a blazing fast commandline search for the NTFS file system. it's very quick but I never used it much....
could it be made useful for FARR by integrating its searching techniques?
http://ndff.hotbox.ru/en/index.html

NDFF is command-line utility that performs extra fast file searching on local NTFS partition. The main advantage over any other file searching tool currently available (including the one found in Windows Explorer) that it looks for file directly in MFT, which is system data table of NTFS. This approach lets avoid reading all the directories contained on the volume, thus speeding up the operation by many times. For example, my 30 GB large partition full of files and directories is scanned in less than 10 seconds, while it takes several minutes to scan it in traditional way.

The main disadvantage of the approach is that the tool cannot look for the file in directory subtree. NTFS partition is the only scope understood by it. Of course, the utility could have filtered out the files that are not located in subtree of interest, but this did not decrease the time to scan the scope wanted. In other words, file search time depends only on number of files located on partition, and cannot be decreased by any hint from user.

This makes the utility useful for searching the files that user does not have an idea where they could have been placed of. On the other hand, it is not efficient to find the files that are known to be located in a dozen or two of subdirectories.