You're right - especially about 4 - but the absoluteness is in practice: your solutions to 2 and 4 require a critical mass of people to understand the data bein

You're right - especially about 4 - but the absoluteness is in practice: your solutions to 2 and 4 require a critical mass of people to understand the data being modelled (and thus - or perhaps the other way around - the processes creating/manipulating the data) and this is so rare as to be effectively non-existent.