Happened to stumble on this whilst searching for TPI® NEXT information. I think the key point is most organisations stop once they get to the low to middle stages of models (the 'defined' or 'repeatable' level and miss the point of getting to the higher levels ('optimising') which is where the most benefit is achieved. At that point, the clever analogy that Michael uses starts to come true.
Continuing on the same theme, it's as if the organisation has gone to school, learnt their times tables, but haven't grown up. It's probably a fault of most models and most organisations using them (I know I've been guilty of this) that once the teenage level of maturity (strict process adherence) are reached, then 'mission accomplished'.
There really needs to be greater emphasis on models and consultants to push to get to the truly high levels of maturity; where the organisation 'thinks' on how to approach different projects and processes are adaptable and chosen according to the conditions and goals. Scott Duncan's comment alsostruck a chord.