For the first part: it depends on what you are trying to convey and to whom you are trying to convey it to. Until the advent of e-mail, text based communication was understood to be a function of preserving thoughts and ideas and communicating those thoughts and ideas both to people the writer knows (as in the case of a letter) or to people whom the writer may not know (as in books, official documents, etc.). In both cases, the understanding was that instantaneous communication was not possible or not desirable (as in the case of a legal document; the directives of legal documents are only enforceable if people can refer back to them whenever they wish, instead of having had to have been present when it was agreed upon in the first place).

Also, in English (as in many writing traditions), writing was generally developed to keep track of things and those that kept track of things were people who were in power, like religious leaders and kings. Those people had titles and power and the language reflects that. It's of great help to priests to have a standard text to refer back to, to base their religious services from, for example.

Spoken language (as far as linguists understand) was developed for coordinating efforts and sharing information, most likely about food and immediate danger. Thus, being understood and understood well by peer groups is of primary concern. Note how groups of good friends who have known each other for a significant amount of time will most often lapse into a kind of shorthand with each other and be able to read each other much better and with less effort than a person that a given member of that group, might work with.

You are right that language is more than vocabulary, but also includes facial expressions, tones of voice, body language, etc. Language learners are often surprised by these mostly non-verbal things. For example, in the US, the gesture for "come here" is done with the palm up while the index finger curls toward the palm. In many parts of the world, this is extremely condescending and is used to beckon animals and only very young children. Instead, the polite version of the gesture in those parts of the world might be done with the palm facing the ground and all the fingers fold down at once from the large knuckles and the wrist might even be used to accentuate the action a bit. In any event, these unspoken things give us much more context in which to hang the content of what is being spoken.

In writing, we only have the context of prior experience with the writer (of which there may be little or none) and the language that is used by the writer to convey the message. As the brain is inclined to look for patterns and meaning, the less context we are given in regard to a piece of writing, the more we will likely "read into" the language of it and the possible intent.

So, long story short, writing and speaking are different. They developed at different times in human history and for different purposes. In both cases, the desire is to be understood and understanding is much easier between people who know each other very well. The better we know people, the less obvious our efforts to be understood may generally be (to an outsider) whereas the less we know people, the more likely we are to use highly structured and detailed language in an effort to ensure understanding.