As noted by LenZ in Ronald's blog, we did apply, but did not get selected. In detail, what happened was that Colin Charles from Monty Program was in contact with his former collagues at Oracle, and found out that they had no plans to apply for GSoC this year. I of course have little insight into the motivations and prioritizations of Oracle managers, but I guess it is safe to assume that "the person who used to be the MySQL administrator for GSoC does not work for MySQL/Oracle anymore" must have been a strong contributing factor. Since Colin is that person, it was agreed that he would submit the GSoC application in usual fashion, after which MySQL developers from any fork or employer could become mentors under that umbrella.
Since we cannot use the project name "MySQL" that is owned and trademarked by Oracle, the application was submitted as "MariaDB" but the text made notice that MySQL's community team is on board with it.
Alas, the GSoC is becoming increasingly popular, and Google's team had a hard time selecting among the many applications. In the end our application was not selected. Since MySQL is one of the leading Open Source projects in the world, and had participated every year, this was a bit of a disappointment. Reviewing the application I can only conclude that on Google's end (even with the personal connections we have and everything) our application may have looked like some small fork, and it may have been unclear to them what the relationship this really has "with the real MySQL", etc.
This is the cost of a project forking. (Or, as Philip Antoniades argued, in this case it's not even a proper fork, yet we carry some of the same downsides here...) In the short term it is costly for each of the branches and things like this happen. It is unfortunate, and one can only hope that in the long term the forking proves worthwile.
Which is to say, better luck next time!
(You can still contribute code to MariaDB during the Summer, we'll be happy to mentor you :-)