I lived in Portsmouth, OH in the late seventies and early eighties. Unlike many larger metro areas, Portsmouth is cramped by geological features (two large rivers and tall hills, some would say small mountains) and thus most traffic uses US 23 (N-S) and US 52 (E-W).
Until about 1984 or so, the lights on both main drags were synchronized so that traffic could flow continuously at around 32-33 mph, just under the 35 mph speed limit. It worked incredibly well on these “pipeline” roads, especially where the traffic was separated, e.g. most of US 52 had the eastbound traffic on a different street from the westbound, separated by a short city block.
When city planners (a phrase generally believed to be an oxymoron in P-town, emphasis on the moron part) activated a “high-tech” computer-controlled traffic system in the mid-eighties, the traffic pattern went right down the toilet, literally with the flick of a switch.
I guess my point is that timing, or synchronizing, traffic lights can be very effective under the right circumstances. In larger cities with less-constrained traffic, however, managing traffic flow is much more problematic, with numerous arteries criss-crossing and buses and cabs entering the mix.
As for expense, simple synchronization of lights on arterial streets is relatively cheap, compared to the cost of programming and maintenance of fancier systems.
Finally, NO system of controlling traffic lights will compensate for poor planning. One of the best examples I can think of is Polaris Parkway on the north side of Columbus. When the road was built just a few years ago, not much of anybody used it because it wound through corn and soybean fields and didn’t really go where anyone wanted to go. Now, after nearly a decade of unimpeded – and largely unplanned and uncoordinated – residential and commercial development, Polaris IS where hundreds of thousands of people want to go every day. Last year, I lived in an apartment just off the parkway, and I had to travel 1.6 miles east to get to I-71; during certain times of the day, that trek can easily involve 30 minutes or more of sitting in wall-to-wall, creeping, bumper-to-bumper traffic. This despite an ongoing program of adding computer-controlled traffic lights. But since the backed-up I-71 on-ramp is the real bottleneck, the lights have little effect, and I doubt if the traffic pattern would change much if they were all removed. Planners (if there were any) failed to consider the effect of all that development, and the associated massive increase in traffic volume, on I-71 itself, which was already overloaded. Similarly, formerly rural county roads in the vicinity, which now have become feeders to new residential mega-developments, have become clogged with commuters who sit in line for fifteen or twenty minutes at every four-way stop intersection.
Unregulated suburban sprawl can be an ugly thing, and no method of controlling traffic lights is a sure cure.