Well, I wouldn't say the Saffir-Simpson scale is *subjective*, but it is somewhat *arbitrary*. A lot of the problem is that we talk about hurricane classes as if they were purely descriptive of storms, but really the various classes characterize the potential interaction of a storm with man-made structures. Each hurricane class represents the likely level of damage to a wood frame structure built with construction techniques common in the US in 1971. This could easily have been calibrated using historical insurance statistics, at least for the more common category 1-3 hurricanes. So the scale is not *subjective*, it's *contrived for a particular use*.
The National Hurricane Center originally took Saffir's wind speed scale and factored in potential storm surge, which I think makes sense. A few years ago NHC changed the scale to be a pure wind speed based scale, which might be more useful for some purposes but I think reduces the relevance of the scale to most people. Take Sandy, which was "only" a category 2 hurricane; the damage was caused by storm surge. Another factor that would go into a really useful scale of hurricane power would be geographic extent. Sandy, while not packing intense winds, was *huge*; that meant that it was going to find the right conditions of pressure, wind and tide to cause major damage *somewhere*. Consequently Sandy caused more damage than any other Atlantic hurricane excepting Katrina -- $65 bln. The next hurricane down the list by damage is lest than half that (Ike, at 29.5 bln).
So the issue of whether we need another, higher level category is not a matter of wind strength alone, whether or not that strength is increasing. It's a matter of needing to characterize a potential for damage to things other than wood-frame structures.