I'm guessing this partially depends on region, and is of course partially dependent on both, but what is the primary factor limiting the lifespan of road surfaces: usage or weather (rain, snow, winter salt, etc.)? Certainly, there are some types of city infrastructure that deteriorate less from usage and more so from exposure to the elements over time (telephone poles, bridges, drainage systems). It seems that increased housing density would increase the amount of tax dollars for a given area of land without necessarily increasing the need for more roads or roadway area. One might argue that higher housing density means more cars, but if doubling the number of cars doesn't half the lifespan of the roads, then it's a win. Also, double the housing density doesn't necessarily mean twice the cars, because higher density regions are generally more walkable and there are greater opportunities for carpooling/ride-sharing/public transit.

So, basically, increase housing density so that more people are sharing established infrastructure. Fund the maintenance of this infrastructure using taxes that are population (or perhaps income) dependent, rather than property taxes.

Oh, and based on the cost and longevity numbers provided in the article, why would anyone ever use asphalt over concrete for a roadway surface? I assume some key factor has been left out of this comparison.