I had wondered the same thing when I first came across this study. Apparently, the same researchers did another study where they repeated the same cycle 3 times (for a total of 12-weeks).
Here is an article Joe Friel wrote on the topic, and it also contains a link to another post with his take on the earlier (4-week, single cycle) study: http://www.joefrielsblog.com/2013/04/an ... study.html
Block periodization seems pretty clearly superior on the surface, and I experimented with a program of it last season, due to a limited amount of training time that lent itself to that structure. I can't claim to have any real quantifiable data, but I did see season long progress with monthly improvements in 5-7min (the only real metric I was tracking) output.
One commenter on the Friel article does raise a good point though, which is that this sort of block structure will create a natural "taper" of sorts, so it could simply be that the improved performance at the end of the trials was not from improved "fitness" but instead was due to better "form" due to being more well rested at the time of the final test. That seems plausible too. To rule that out, it would be interesting to see another 12-week study where they had a longer "washout" period of identical training between both groups leading up to the final test. This would hopefully preserve their respective levels of fitness but allow them to be equally well rested at the time of the final test.