There's another interesting quirk I noticed, but as MS isn't going to pay me for my research, they can get billed by 'professionals' who also figure it out...
Of course, the file may be corrupted even if it does get downloaded. I can tell Internet Explorer to download it again. "File exists - replace?" it asks. "Yes." Does it replace it? No! - it checks to see if the file appears to be on the disk, and it then pretends to download it. But in fact, the "download" takes place in a fraction of a second, and the same, corrupt file is left on the disk. The only way of getting the correct file is to go to the disk directly, delete the corrupt file, and then go back and download. Again, Redmond knows what is best, and my opinions, as the operator of the machine, can be safely disregarded.
Not only do you have to delete the corrupt file, but as explorer saved a temp copy of the file even before it saved the 'real' copy of the file, you need to find the temp copy and delete it also. It usually is named some gibberish like 'dhdsrsfxvxdvksdfkZCZASf.tmp' and unless you know the filesize or what your looking for the only way to delete it is to go into IE's properties and delete all offline content and restart IE. Otherwise IE will fake the download and copy the temp file back into the 'real' copy of the download- thus preserving the corruption.
Thank you Microsoft, may I have another ?