Well, the answers to all of the questions is that a particular grain is used for a particular purpose. The cross section (roughly diameter to light) is the main thing you want to consider. Second is halide content, and then the dispersity (variance in grain size / unit grain count) and on and on. So, cubes may be as good as octahedra depending on purpose.
T-grains mainly have the advantage in cross section. They have a huge diameter wrt light per unit volume as they are very thin. So, you can stack a lot of them and get fine grain and high speed both (in a sense).
Grain shape is determined by starting vAg and growth vAg. I have published that chart and discussed it in another thread, so I didn't repost it here with the electron micrograph. You may make a cube and yet end up with what I showed above simply as a function of digestion and ripending. The ripening dissolves fine crystals and deposits the AgX on the larger crystals growing them even larger and 'softening' the outlines. Digestion directly acts on the crystals as a whole, dissolving them and reforming them as the process continues. An ammonia digest then, carried to completion will yield what looks like a soccer ball by virtue of changing all crystals into one type.
The aim, whatever shape, is to get either a uniformly polydisperse emulsion, or a very monodisperse emulsions. In the first case, it could be used directly as a single component film emulsion with rather low contrast, and in the latter case it would be used in a paper. A monodispersed emulsion blended with several different emulsions will give a low contrast film emulsion. Thus a monodisperse 800, 400 and 200 emulsion set would each be too high in contrast for a film but blended will give a good low contrast film. This is true of any crystal habit.
All of the above is an approximate generalization.
I found a page on the Harman site with images of crystal growth and small descriptions. Its very interesting.
The coating page
There is an Agfa site that shows a slide show of the cascade coater in action. This, the addition of active chemicals, and making exotic emulsions, is presently out of reach of the home darkroom worker.
I am approaching this in several stages:
1. What can be done in an inexpensive home darkroom.
2. What can be done by an advanced emulsion maker.
3. What is the most that can be expected in the home darkroom making emulsions.
I have dabbled in all 3, but am presently stuck at 1.0 for the benefit of the average user. The picture above is beyond the average user in availability and expense, I would think. So, it falls into 3. I just wanted to show what is possible using the methods in 1.