Just kidding, make whatever game you want.
For me throwback means revisiting some of the quirky old innovative DOS games that I used to play as a kid.
There were very few platformers then.
It's kind of sad you feel that way, as there are devs such as Managore that proved more than once we could push the boundaries of what is possible to do with platformers. All with jam games like WindowFrame, Birdsong and Roguelight. Not to mention, the professional devs out there that created hits with this genre such as FEZ and Braid. True, we have a flood of platformers that don't add anything new to the table, I myself am guilty of that. But a platformer is just a design mechanic that influences how you implement ideas, just like the engine or the programming language you use to make games. You just have to be experienced and know how to be inventive with it. Now let's make some great games!
I absolutely love Daniel Linssen's works, and most of my all-time favourite games are patformers (Cave Story, Braid, FEZ, VVVVVV). Of course it's totally OK to create a platformer, but when there are so many awesome games of the same genre, it's not enough to make a decent game, you have to make a really good one, an interesting one, like Daniel does with his platformers.
The problem is not with the genre, for me personally, there is no problem at all. But if one wants to create a platformer, one must try really hard not to create "yet another platformer", since the underlying mechanics are so frequently used that it got pretty boring for a gamer to just jump around killing enemies while trying to follow a somewhat cohesive story.