There are hundreds of thousands, if not millions, of infected home systems trying to deliver spam. I used to examine my mail logs to look for patterns, but there are very few duplicate IP addresses. And even if one system did try to deliver enough to trigger such a filter it would quickly be replaced by another. Eventually you would be blocking huge portions of the 'net, one IP address at a time. You could get more sophisticated, and maintain a database and try to consolidate by netblocks, but the end result is that you'll probably just block most of the Internet.
And a significant number of legitimate sites, too. :-/ I personally use that approach. I don't accept email from sites without valid rDNS unless they've been explicitly whitelisted. There have been some important emails blocked because of it, but I say, "too bad." I've tried to inform the site admins, but they usually ignore me. One company has multiple mail servers, and some of them have valid rDNS while others don't. So random emails from them bounce.
Unfortunately, as the namespace becomes more crowded it becomes more likely that previously invalid names will become valid ones. You could, of course, make them obviously invalid, but if they're obvious then they're easy for the harvesters to filter.
John Meissen snipped-for-privacy@aracnet.com