I think that the article is almost too vague to comment on. There is a statement about insurance companies suggesting that architects never advise on means and methods, but it doesn't describe why that wasn't a defensible tact.
Tony, you ask, "How can TAP best engage in creating a more reliable process?"
While many want to sing the praise of BIM tools, I see their utopian reliance on perfection to be the stumbling block that keeps them from being truly useful as an efficient means of project delivery. Who, besides myself, has been frustrated to the nth degree by error messages while building a model? (Of course, the irony here is that the ENR article is about construction sequencing. BIM products are anything but intuitive, and require precise sequencing to prevent the design process from stalling.)
I am not interested in a 100% perfect model. What would be really useful is a model that conveys the appropriate information and gives you an idea of how imperfect it is. For example, a summary report of the model could include information like, "97% perfect - potential errors in wall construction joinery", or something similar. There is plenty of room for interpretation, but assuming that any drafting that an architect does is a 100% perfect reflection of what a built building will be is pure hubris.
Software companies are the dog, and, unfortunately, architects are the dog's tail. Also, most unfortunately, a dog's tail is very proximate to another part of its anatomy.
-------------------------------------------
Ken Brogno AIA
Architect
AIA, LEED AP
San Francisco CA
-------------------------------------------