First of all, who in tarnation is this Mike Flacklestein character? Search for the name, you find thousands of results, all apparently similar blog comments asking about random addresses, all on the (as far as I can tell) nonexistent Commonwealth Street in Seattle. I don't appear to be the first to have noticed this, but I can't uncover any sensible explanation.
However, a few ideas come to mind, if you're as devious a fellow as myself.
When I saw the first one I assumed it was a real comment, but since I've never been to Seattle and didn't recognize the name I gave it no mind. The second one got my attention though, which led me to do some searches. Looks automated, then, given that generally only the number changes. Normally that would indicate comment spammers, of course -- miscreants who post links to their webpages in bulk, buried on old blog posts where only robots will find them, in the hopes of pushing up their ranking in the search results. This generally doesn't work, because these days most everyone uses the nofollow tag, which marks untrustworthy links such as those in blog comments for search engine spiders to ignore. But it doesn't cost the spammers much to try.
Except in the case of Mr. F here, there are no links. He has a Gmail address, and lists Google as his homepage. Somehow I doubt that Google has to resort to comment spam to improve its positioning in the search engine world. Search for Mike Flacklestein; all you'll find is these meaningless comments -- and, if a few days, probably this post. So what's up, if it isn't spam?
Maybe it's just a prank. Someone thought it'd be fun to spread a meaningless name across the globe. The two comments I have come from separate DSL accounts in Irvine, CA, judging by the IP addresses. But two is an awfully small sample, so I wouldn't draw any conclusions from that. I'd like to know where others originate. But there are potentially ways to exploit something like this to more interesting effect.
Observe: each Mr. F comment is identical, except for an apparently random five-digit address. That's log(99999)/log(2) = 16.6 bits of information per comment. The messages have long common segments, so they're easy to search for. Google reveals at least hundreds of these things, so that adds up to perhaps a few kilobytes of information, easily and anonymously retrieved with a web query that leaves no trace as to who posted the information nor who collected it. However, using this as a bulk data transfer wouldn't be very efficient. You'd have to be able to reassemble the message; uniquely identifying each of several hundred would take 9 to 10 bits of your 16.6, but that sill leaves a couple of kilobytes of compressed, encrypted data to play with.
How about posting the comments, then? Surely that would lead back to the sender, if a few blog operators got together and compared IP address logs? Not necessarily. Most comment spam, like email spam and more malicious activity, originates not from the well-hidden Spam-Cave but from widely dispered networks of desktop computers running unpatched versions of Windows on DSL lines and corporate networks. Bored teenagers can download tools that automatically scan the internet for such vulnerable machines, hack in and infect them in seconds, and thus render them a horde of zombie computers able to wake up at a moment's notice to do the bidding of the, well, highest bidder. Often without incriminating fingerprints that might lead back to said bored kid, and almost always without a whiff of a trace of the purchaser. Posting a few thousand strange blog comments from random computers across the country -- this is trivial. The right "script kiddie" would do that for free as an introductory offer.
Which led me to another notion. Some zombie networks have been built not by kids with scanners, but organically by self-propogating worms. (Remember those worms that occasionally shut down the Internet a few years back when an infection got out of hand? Now they've mostly been tamed, and put to work.) Trouble is, if you want to control such a thing, each computer has got to phone home somehow. Make them all call some master computer, you might as well light up a Bat Signal from your roof, 'cause that's getting shut down. More common is for each computer to pass messages back along the chain of infection to the source, but if the chain breaks large portions of the network can be lost. But what if a worm, upon arriving in a new host, instead connected to the internet behind the scenes. It would be easy to find a blog at random and leave a comment. At 16.6 bits per message, only two such comments would be needed by each worm to broadcast its 32-bit IP address, and thus reveal its position to whomever's pulling the strings.
In either case this would represent a form of Steganography, the trick of hiding information in plain sight. Not especially good steganography, of course, since after just a few hundred or thousand postings it's already readily apparent to someone like me. That doesn't especially matter if the included secrets are encrypted in some sensible fashion, though. Without the ability to read the message, discover its origin or destination, or even hinder its delivery appreciably (by deleting the comments on my own blog I could only destroy a fraction of a percent of the message, and any scheme like this would have to have built-in redunancy) there's little sense in which my knowledge of the existence of this channel of communications can credibly threaten its users' designs.
On the other hand, maybe it's all just an ARG about a sentient AI loose on the internet that's trying to make contact. It's not like that's ever been done before.