This is so hilarious that it feels wrong to nitpick it. But in the interests of science, this survey technique:
A scan of Twitter reveals a lot more people talking about actually using Match.com than Plentyoffish...
... is like a freshman-class exercise in Spot The Sample Bias. Twitter users are hardly a representative sample of the online population, let alone the population at large.
Twitter users alone is a small sample, but I think there are other anecdotal samples that support (but can't prove) the author's hypothesis. I can't remember anyone talking about PlentyOfFish, and this includes people who have tried multiple dating sites. Compete, Alexa, Google Trends and other traffic estimators also show Match having higher traffic. The only statistics showing PlentyOfFish as bigger than Match are the comScore numbers quoted by Mashable.
It's also important to remember that Markus Frind was accused of overstating the PlentyOfFish traffic levels to generate buzz years ago when the site first became popular. Although the post went off into a satirical conspiracy, there could still be something fishy going on.
Well, it could certainly be that Frind's stats are off -- accidentally, or on purpose, or for the same reason that many scientists' stats are off. (They make an initial mistake, and the positive feedback from that mistake creates a big disincentive to reexamine the data too closely.)
If so, it's a joke on the web stats industry, which seems to be... well, a joke. If people had any confidence in the accuracy of these traffic estimators, there wouldn't be room for a wacky conspiracy theory, right? Does anybody really know how many people visit a site, besides (perhaps) its owner? I suppose Google might give me more definitive data if I paid enough and signed an NDA in blood.
A scan of Twitter reveals a lot more people talking about actually using Match.com than Plentyoffish...
... is like a freshman-class exercise in Spot The Sample Bias. Twitter users are hardly a representative sample of the online population, let alone the population at large.