my 2 cents:
The article is respecting the LAMP space while providing a basis for Openacs. But pitch it against a similar article by a LAMP solution provider. I feel that we might see the basis for quite a few points eroded like the multitiered architecture, multithreading , data model ( depends upon who our clients are). I personally have great respect for openacs datamodel and overall architecture in general but when coming to clients who are concerned whether they have x number of features , stability and scalability in performance , a data model argument may not be up in the list.
IMHO one of the great strengths of Openacs i.e rapid web application development can be emphasised more.