It doesn't take into account future increases in BTC value, either (which could offset this). There are many variables, but I for one think this is a useful tool for demonstrating that mining can be profitable.
Of course it is profitable (certainly if you already own hardware) but calculations without factoring difficulty increase are very misleading.
BTC price increases are possible but definitely not certain. Heck, the BTC price can even fall hard. If there was a futures market, it would show very similar prices to the spot prices because future is not know and the current value is the best bet. But as long as mining is marginally (i.e. BTC income larger than electricity costs) profitable, people will join mining and difficulty
will increase (that's how arbitrage works). And if the BTC price increases, the difficulty will increase even more. The rate of difficulty increase is not known but this thread is a proof that people still join the mining game and we are far from difficulty saturation.
I don't think we will see difficulty increases of 20% per 2106 blocks (as mrjones suggests) but 10-15% is very real in the nearest future.