1.) Is the vertical resolution more important than the horizontal resolution? If so, Why?
2.) Is it better to encode the full resolution from the beginning, or stuff it into a smaller resolution and then let the decoder stretch the image? What happens to the bit/pixel ratio after stretching?

So, which one would be better: 624x416 or 544x480
576x384 or 480x480
496x320 or 352x480