Shorthand summaries of the paper's argument note that firm managers face many "make or buy" decisions. Buying involves transactions costs while making involves monitoring costs. Management is hard work because these are tough choices that must often be revisited and reevaluated.
Another way to say it is to ask how much of any product's supply chain should be within the firm and how much should be beyond the firm. Here is a fine discussion of global supply chains in the modern world.
All this brings up the corollary of organization. Make or buy? If buy, then buy where? Locally or not? Needless to say, the way our cities look hinges on this elaboration of Coase's question. How much proximate co-location do we get? Consider these two views:
“ … a central paradox of our times is that in cities, industrial agglomerations remain remarkably vital despite ever easier movement of goods and knowledge over space.” “Introduction” in Agglomeration Economics, E.L.Glaeser); and
“When co-location is infeasible, networks may substitute for agglomeration. This possibility of substitution means that small regions may survive and prosper …” “Agglomeration and networks in spatial economies” B. Johansson and J.M. Quigley.
I do not think there is a paradox. I do not think that it is either-or. The agglomeration vs networking choice is linked to the make-or-buy choice. It has long been recognized that location and trade are part of the same puzzle. It is a huge and complex puzzle that Coase properly saw as way beyond "chalkboard economics."