The paper seems to find a repeated issue with using areas where there are high density of buttons. The small area makes selecting the correct button difficult and frustrating. The issue of incedental touches remains but is being addressed by the ability to "cancel" a touch. Touchscreen typing is still slow and disliked. Small tap areas also cause frustration among users. Multiple swipe areas in close proximity can cause significant frustration.
Smart Phone Usability
The authors have found that the primary source of user frustration and dissatisfaction stems from poor software as opposed to hardware when it comes to smart phones. The paper is somewhat dated so I would be curious what their findings would be if they were to do this today. There was no review of iOS or Android so the findings are somewhat dated and non applicable as each of the OSes used are now considered obsolete. The principles still do apply to today's phones. Navigability should be fast and easy, especially when working on a device which can be as frustrating to use as a smart phone.
Although I share the same sentiments with the author regarding new UI for productive work, I disagree with his belief that gestural interfaces don't have their place. The author seems to think that testing UI design should be secluded to the lab but I disagree. The reason in the past why Lab testing was so critical was that shipping an unfinished project or untested product can result in failure. But with the pace of current technology, it is more costly to keep a product in localized testing to prevent a potential flop than to not ship the product and fall behind. The author seems to put undue criticism on Apple/Google for the pitfalls of the developers. I will agree that the two companies should have probably given better guidance to the developers as to how the UI should perform, but the blame is primarily on the developer.