The writer was suggesting that real R&D costs aren't anywhere near as high as the pharmas want us to believe.

It's easy to assemble anecdote to "prove" whatever viewpoint people want. for example, at the moment some lobbyists claim that ambulance-chasing adds only 2% to the cost of healthcare. In that case, why does US malpractice insurance cost six-figure sums- 10000% higher than elsewhere in the world- and why do US physicians openly admit that they over-order to deny lawyers any avenue to criticize?

In the case of pharmaceuticals, R&D proceeds in multiple directions in the hope of finding a single success that will justify the entire effort. If you calculate the R&D component only of the successes, it may seem small. Add in the real costs of the entire investigative process and the costs borne by pharma are staggering. Obviously they need bumper returns to justify the risks they take, investing billions of dollars trying to find something useful. Remove the benefit and it won't happen- the top people are better off going to work for Goldman Sachs where they can earn millions doing nothing particularly useful.