Just to amplify on the above, going also to the basics of digital video.
1. The optimal time length of the clip you want to shoot depends on what is going on. If not much, then the viewer's mind will wander. If there is a lot going on, then the clip length depends on the interest and scope of what you are shooting. A shot of a baseball player hitting the ball, running to first and rounding the bag takes more than five seconds. A shot of flower in a pot that goes on and on is boring. Get it? On TV, it is mostly close-ups of heads talking. So the viewer is not visually bored, they usually shoot in short clips, changing the angle while the actor pedantically describes some plot point, or carries out a fascinating test tube analysis of a fluid sample.
2. Video files are sets of zero's and ones. That's it. There can be errors in transcribing zero's and ones, but there can be no difference in "quality." You are thinking analog. Everything is numbers. In the old days with analog TV, you could adjust the antenna to get a better picture - less noise. With HD TV, which is all zero's and ones (it's digital), you can either receive the signal or not, adjusting the antenna does not give you higher quality. Similarly, the sd card can either keep up with the bitrate (see below) - how fast the zero's and ones are written - or it cannot. It will just fail if it cannot, no change in quality.
3. Again, video files are just bits. The size of the file depends on the - you got it, the *bitrate* and the amount of time you record. That's all. The video can be HD or SD, or high-frame rate or low-frame rate, any kind of compression scheme, any shutter speed. But the only thing that matters for size for a video of a given time length is - the bitrate. How many bits are being written per second. So if you shoot a 120 fps clip for 5 seconds or a 60 fps clip for 5 seconds and the bitrate is the same, the file sizes are identical.