Hi,
thanks for your very long feedback. Please allow me to go into some of your points.
Regarding NTFS / FAT32, you just got the wording of Microsoft mixed up. They see NTFS as superior to NTFS and claim that FAT32 can only handle 2GB files (which for Windows is true). FAT32 partition size is artificially limited to 32GB by Microsoft (on Linux, you can create FAT32 partitions a lot longer). The main reason we are sticking with NTFS at the moment is that it offers several technologies we require for file tracking, which include additional metadata per file, for instance.
CAPS vs no-caps: thank you for your input. It sounds like an intuitive approach.
Auto-Startup: using or not using auto-startup of tag2finds frontend application is a matter of taste, if you like it or if you do not like it. It has no negative impact on the functionality to start tag2find on demand. It is, however, vitally important that the tag2find service is set to auto-start, as without this service file tracking will be basically disabled and tags might be lost. The service is able to compensate a certain down-time (which is as long as the NTFS Change Journal can keep records), but it will try to catch up as soon as started which might take considerable time with high CPU usage if it was not consistently running, or, if entries from the NTFS Change Journal have already expired, it might even cause lost tags or our database and the filesystem to be out-of-sync.
Auto-Update: we have an new-version notification which is enabled by default. It will tell you in a so-called "Balloon Tool Tip" on the notify icon (next to your clock) that a new version is available once per day. It will also tell you that there is an update available in the TagBrowser status bar. We want to introduce a more-featured auto-update later, which will automatically download and update your version of tag2find if required.
Extended Searches: we are already working on this, especially the mentioned "-" operator for excluded tags. There are a lot of use-cases which require this and we are sure to provide this.
"Untagged": this is a great idea, it has some minor implementation consequeneces though, as we'd need to maintain an index of all files, not just the ones tagged on the file system. But you are definitely right, this would be very useful and we will try to find an efficient way of handling this.
File-Search: again I just can tell you we are working on something similar like this already. I'm really happy that you show us that we are thinking along the same lines as you, our users, for whom we are working in the end.

Thanks.
Again, thanks for taking the time to write such a long post. We really appreciate any feedback! Hope to hear from you again.
Best regards,
Martin