Sorry, CP, but I'd actually not recommend starting like this. At least, that's not how I start! I guess that everyone has their own method for starting songwriting, so I'll just share mine.
The most important part of the song—to me, at least—is the inspiration for it. You need a good idea for a song for it to be worth doing. Nice thing about the fact that smartphones have become nearly omnipresent is that you have a microphone on you at all times. If at any point during your day, at any time, you find the inspiration for a song in the form of a melodic nucleus, or a rhythm seed, or anything like that, record it.
I find that inspiration is a fickle, effervescent, ADHD thing that will escape your grasp if you don't put it somewhere first. People will look at you very strangely for singing into a phone while on the subway, but if it means that you have your idea somewhere other than your ADHD brain, then that's a good thing. And besides, if when you get home, you find that your idea sounds awful, it probably wasn't a great idea to begin with!
After that, expand on your nucleus a bit, and reflect on it. A melody alone does not a song make. What would complement your song? Is it original? Why are you writing this song? Only when you're sure about these questions do you continue with composing your song proper.
The reason why I don't start the way that it seems CP does is because the things that she mentions seem like semantics and ancillary information. Generally, when I start with a song nucleus and build off of that, the genre, tempo, time signature, and instruments follow. (The key signature doesn't matter a whole lot. Pick C, F or Am if you don't like sharps/flats in your key signature, though.) If your melody is in 4/4, so be it. If your melody is in 6/8, so be it. If it sounds good, and you're willing to do it, then it's a good idea.
No offense, Coco. Just ringing in with what works in my own experience!