If it didn't recognize them as distinct URLs, a significant percentage of links would be rejected as having already been submitted. Http://site.com/article.php?id=1234 and http://site.com/article.php?id=5678 may be two legitimately unique articles.
My point is that it's not exactly rocket science to add an if/case statement in the Submit script that checks whether the URL being submitted is on a list of domains for which the '?' and anything after it should be disregarded.
'Twould also be great if it recognised that www.bbc.co.uk and www.bbc.com are effectively the same site.