To keep a long story short, most polytonic Greek is posted in precomposed form and displays well in IE7. But sometimes Greek text has been produced using combining diacritics. It is a pain because that makes two standards and you can't tell with the naked eye which is which. At least if they display properly you shouldn't be able to. But you can see in the first image how they both look in IE7.
The second image shows both kinds of text in their correct display form in New Athena Unicode and they should look identical.(Are you bored yet?)
There are many pros and cons to all this, but I believe that the usual way to display Greek is by precomposed text. This is the opposite to Hebrew.
As far as Polytonic Greek is concerned, the best resource is Rodney Decker's resources. This page will take you from the age of the dinosaurs up to the present, so I would recommend starting to read this paper on page 10 or 11. The relevant stuff is on pages 14 and 15. Read point 4.3.3 and on footnote 3 on page 15. In fact, if you have to decide in any paper which part to read, I would recommend the footnotes.
In short, most polytonic Greek text is produced in precomposed text. Zhubert produces precomposed as well, so no problem. This is what the MS Greek keyboard does too and probably most of the others. But sometimes there is a bit of the combining diacritics stuff around. For some reason IE does not handle this well, although Firefox, as usual, has no problem. Go figure. BTW, here is a MAC biblioblog.
Anyway, IE7 now does font substitution like Firefox so fonts are not the problem - codepoints are. I use Babelpad to look at things like this along with Babelmap, from Babelstone, along with reading Babelstone Blog, which is a pretty cool blog about Unicode. Except that Andrew is writing about the Morrison Collection right now, so you have to check his archives.