Hmm, no I haven't actually noticed problems that I can recall with incorrect characters and I did notice the pages were encoded us-ascii, but I also noticed the news-feeds themselves when subscribed to have encoding=utf-8 in the header and was a little confused. I simply defaulted to ISO-8859-15 out of habit.
So anyway, as you say it shouldn't matter as all the characters in the articles do seem to be entitised, thanks for clearing that up oneillpt
