Wow, while I have been off experimenting, this has turned into a real thread! Thanks for the interesting input. There would not have been much point asking for advice if I wasn't going to follow it: so I tested and tried NTFS with the intention of using it.
I managed to reproduce silent data corruption under normal usage scenarios with Linux NTFS drivers that you may be using.
To put it mildly...
*that is quite frightening*
...so I am have started a new thread about that. It also proved to be a pain in the ass. Some of your questions are answered in the new thread instead.
Linux NTFS data corruption observed
Yes, my Windows is hybernating data-only partitions as well as the Windows OS partition, more info on that in the new thread.
UDF was designed for use on CDs and things like that. Thinking about it, it may actually be a good filesystem for SSD devices in the future. I think most people (me includeded) forgot about UDF when the Rockridge (methinks?) extentions became widely supported on the normal CD filesystem. It is certainly worth pondering now that you mention it. UDF will need to be modernised if they expect people to use it into the future tho'. The maximum partition size seems to be 2TB.
I discovered that NTFS has surprisingly tight restrictions on maximum pathname lengths. It's really not a big deal, but I thought it was interesting that I hit a limit in the real world. I had to rename some of my ext2 stuff to make my data fit on NTFS. Weird - I always thought Microsoft were the ones who liked long filenames!
Next I will be running with ext3 (using the extfsd drivers on Windows) for a while. I will aim to report any problems here. If I post no report here, it means that it is proving to be 'adequate'.