OK...thanks for the clarification.
There are a couple of things to consider. One of those is access to bandwidth. I am going to presume that you have access to whatever you need in that capacity. We can have -that- actual discussion later, but for now, let's presume you can get what you need.
There are many things to consider, so I will give you the "big picture" on how I would proceed. I'm going to paint with a pretty broad bush here...you can adjust accordingly.
You are going to need to have the cooperation of the owners of the "vertical real estate" in that area. Fortunately, most of those folks might be sympathetic to your cause. For example, the Homewood Early Childhood School has some vertical real estate that is good.
I see by looking in your area that there are a number of tall buildings. One of those is the Pittsburgh Student Achievement Center. I am going to presume that you can get the bandwidth you need from them...just for this example.
At that location, you install a 5 GHz radio that transmits to other more local nodes, one of those being located at the Homewood Early Childhood School. At that node, it beams out a 2.4 GHz signal to those in the neighborhood that want to receive it.
Now, how your potential subscribers receive that signal is in their hands. They might have a smartphone that can take advantage of the wifi signal, or they might need to purchase a low-cost ($55) adapter to receive the signal at their house
Let's stop at this point and look at what we have...
You are providing the backbone to a system that serves many people. At this point you can do that since your one system can provide service to about 120 people (two nodes with three channels with 20 connected people per channel). Your cost at this point is small, since you are only providing the main radios to service the many connected subscribers.
If you have to add the cost of providing all of their radios, your capital expenditures escalate rapidly. The "sweet spot" is to prove the backbone, and let the user pay for their own network connection device. In this example, there is a "buy-in" cost of ~$55 to get the connection device, and the subsequent monthly charges ($10 in your case) to subscribe to the service.
The upside is that most new phones already have that capability, so your subscriber has no additional costs if they are using that device.
There also needs to be a billing and accounting system of some type. There are low-cost standalone devices that take care of all of that. This is something that needs to be done, even if there is "no charge" for the service. Network management practices demand it.
So, here's what we've got so far...
The Internet connection comes into the Pittsburgh Student Achievement Center, where it is routed to the roof, and beamed out to other radios listening in the 5 GHz band. One of those radios is at the Homewood Early Childhood School, where it receives the signal from PSAC and relays it on to the 2.4 GHz channels for local distribution (say 300 max feet from that location).
If you add more 5 GHz nodes, you can add more subscribers. Cost per node is one 5 GHz radio at $100, a switch at $70, and three 2.4 GHz radios at $80 each...$410 plus bits and pieces equals $500 per node, if things are good (they look like they are, based on the sat photos on GE). The billing controller is $300 (and you DO need it) but that is also a one-time cost and you only need one of them.
There is the issue of the upstream bandwidth costs, as previously mentioned, so that needs to be taken into account at some point. You might be able to get it "donated".
In my view, this project could be done...it needs a 'sharper pencil' put to it, but generally speaking, the numbers are reasonably good, given your particular objectives.--
Nothing makes an American want to do something more than telling them they can't.