A full disclaimer is in order. I work for AT&T Research and my specialty is wireless. But here I offer my personal opinion and I do not speak for AT&T.
It is unfortunate, that too many technical people on slashdot, who should know better, still think that wireless cellular network is just another "series of tubes" that carry bits around (senator Stevens, anyone?). In reality wireless communications is very different from wireline (copper, fiber, cable) communications due to the physical nature of Radio Wave propagation in open media (air) versus transmission lines (copper, fiber, etc). It is basic physics and scarcity of useful spectrum that is responsible for most grievances of wireless users rather than "evil telcos". Your suggestion to increase cell tower density does not hold water and here is why:
(1) You cannot simply increase the density of cell towers to solve capacity problems. Modern cellular networks in densely populated areas like NYC are already interference limited. Just imagine a New York Stock exchange floor where every broker is shouting like crazy all the time and the noise is unbearable. Now you double the number of brokers on the floor and make them shout twice as loud. Would your network capacity increase? As a matter of fact, it would actually go down because noone can decode anything.
(2) Radio wave propagation characteristics put fundamental limits on what each user can do. Every bit sent to a user that is far away from his/her serving base-station (high RF path-loss) costs more in radio resources to a carrier than a bit sent to a user close-by because it has to occupy wider bandwidth (OFDM) or it would take more time slots (TDMA) or spreading codes (CDMA). Distant user takes radio resources away from other users and contribute harmful interference back to the system. That is why average user data rate is so different from its peak rate under ideal conditions.
(3) Unlike wireline networks, wireless is a shared medium with much less capacity due to shortage of useful spectrum. You can easily lay out fiber links with lots of excess capacity that nicely cushions peak load. Cellular wireless networks, on the other hand, run at much higher utilization and load levels and suffer from the ugliness of terrestrial radio wave propagation and interference.
There are some Good News though. New wireless technologies are coming. 3GPP LTE (Long Term Evolution) and WiMAX are two similar technologies that will squeeze some more juice from existing airways. But it would require huge capital expenses to roll out new networks. Moreover, IMHO user demand grows still faster that the capacity gains achieved by those technologies. More spectrum will be needed and it is very expensive and scarce.
Someone will have to pay for all this, right?