Well, then let me first be more elaborate on what I want to do. As I said, I already have a well-edited collection, and I think even for new albums, I can get most of the info (artist, composer, etc, or even original album for collections) from my existing library.
Now, no matter whether the tag fields of the new files are empty or not, there is a chance that I would end up with duplicate entries as a result of spelling mistakes, use of initials, etc. E.g. for an artist name like A. R. Rahman, I might end up with AR Rahman, Rahman, A. R Rahman or such slight variations unless I copy paste the exact spelling I use in my existing library. It may not be an issue for many names, but for others it is a big issue. So, there are two problems actually: I have to type the names (as auto-complete won't in Mp3tag without a database) and there might be duplicate entries for the same artist/album etc. The only solution is to copy-paste, as I said. But with too many files and too many fields, it would be a real pain (batch edit may not always be the option, though that's the only relief now for certain albums).
Now here is what I can do with a database, for example of MediaMonkey: no matter which field it is, artist, album, composer, original album, etc. I can simply pick the name from the drop-down list (like with genre field in Mp3tag) or if I type a few letters, auto-complete option completes the rest (esp. with MediaMonkey, auto-complete works with separators too). So, with the simple option of using my existing library, I can bypass all the problems: I won't have to type everything, I need not copy-paste a bit, and the bonus, which is more important, I won't end up with duplicates.
Then, what's the problem with using the media player, you might ask. First, they handle tag editing a bit differently than I am used to with Mp3tag. I end up with multiple entries sometimes (multiple year, artist etc.). Also, I have a lot of duet tracks, and the separators those programs use create multiple lines which makes it impossible to see the latter half from say windows explorer and such, and also, sometimes those fields get omitted if I convert the file.
Also, as you people mentioned Mp3tag is quite good to work with, as I have a lot of flexibility as to what I want to do and how. The only problem is that there is no way I can use my existing database.
One thing though, as I mentioned earlier, if I add previously edited folders to Mp3tag to help with this, that does not work exactly the way it should. E.g. if I have a duet with artists AA and BB, I would have to find another file with those exact artists. If I have them separately in my library, that won't work.
I hope this would help in understanding the exact problem and what I want to achieve. I hoped that others might have faced similar problems and might have some kind of a workaround. Somehow, I could not find any yet, though I do use the 'actions', etc often.
HansBKK, bother explaining what you meant by this?:
"Or just invest a bit of time up-front in setting up a well-automated standard import/export workflow - then you can do your data-entry in whatever tool you like and still use MP3Tag to actually write the tag data."
...and if that would help me with my problem
And, ohrenkino, you are right that I still have to know the file's details, but the above explanation should make clear my exact problem. A database will not only make tag editing faster, it would also be accurate. If I get tags details from elsewhere, for Rahul Dev Burman, I might end up with R. D. Burman, R. D. Barman, RD Burman, R.D.B., etc. etc. Now I know they are all the same, but my computer does not. So I do it my way and try to keep things synchronised, and that is where the idea of a local, manually edited database comes in.
Thanks again for your help, I really appreciate it.