Stop being analitical!

[Emma May Smith]

One observation was that when k and t are prefixed with o, both become more frequent. My explanation for this observation was that this happens since o + d is normally transformed into o + gallow.

The idea that when d is prefixed by o it becomes a gallows is an extraordinary claim! What's your proof?

One factor is therefore  that t is preferred in line initial position and qo- is not. But this is indeed not the only reason. Another reason is that qo- is far more common in Currier B and that most of the qo- words in Currier B are starting with qok-. Words starting with qok- are therefore in general more common then words starting with qot-:

You're simply restating the observation. You can't say that one word is more common than another because it occurs more often. 

Why is qo more common in Currier B? Why is qok more common in Currier B?

Also, what has this got to do with your earlier claim about space and gallows preference? Now you're saying it's not that important? What?

I think that the only places where there is less space is when the text hits a drawing.

With less space I didn't mean that there is not enough space. My point is that the author writes k instead of t in a line since this way overlaps with other characters are less likely. This is therefore a layout question. 

There is no doubt that the character combinations where the characters would overlap are less likely within the VMS. For instance a gallow glyph following l is most likely a k or f and less likely a t or p.  An example for a word using lk is olkchedy and an example for a word using lp is olpchedy. For writing olpchedy extra care is needed to prevent an overlap between l and p. That glyph combinations like tk or lk are rarely used is one of the reasons why the Voynich script looks fluid when writing its "own language".

Why us f and p at all in this situation? Their bent leg is worse, much worse, than the left loop of a t.