Hmm, Im having trouble seeing how this simple model adds insights into knowledge generation or design, which is full of essential aspects.

A few thoughts; where is the diverge and converge aspect of design and collaborative engineering? I.e. isnt it a danger of thinking that knowledge always is compressed in Funnel? Doesnt testing generate new insights and modify algorithms? Algorithms may generate bad data that is false, unusable for people or lead to bad decisions? The same applies for code.