i wish dan gillmor's new moveable type-driven weblog pinged blo.gs directly. it's one of those blogs that weblogs.com frequently gets spurious updates for. i suspect dave's weblogsComHelper tool, which appears to be the source of many spurious updates on weblogs.com. i can understand why he made the tool (blo.gs does some polling for the same reason), i just wish it were better.
it does make me think that the ping interface should perhaps have a flag that people could pass in that said “hey, i'm not really the owner of this blog, but i think it was updated” so that the service could take that into account when deciding if the site was really updated. in the case of blo.gs, i'd probably apply the three-day rule in ignoring such unofficial pings for sites that had previously pinged “officially”.
Do you need another sort of dupe, where you either don't accept pings from the previous address, or possibly evaluate the ping to see whether it's still a dupe? Not only do you have cases like this, where you never want to accept a ping from the old address, you also have the potential for situations where someone moves from a blogspot subdomain to a new address, so you dupe them to the new URL, and then someone else moves into the blogspot subdomain, and their pings get counted as updates for the previous owner. My first instinct would be to look at the blog name in the ping to see if it has changed from what it was at the time of the forward, but then if someone runs a poller that grabs the blog name from the HTML, you would end up thinking it changed hands when the page went from dead to a 404. Hrm.