Have you been doing dense high (meaning 250-500+Mbps) wireless?
Have you built out fiber in a neighborhood?
If not, then I'm guessing you are talking about a different "WISP" scenario, not comparable to what I'm talking about. So Terragraph doesn't apply.
If so, then you should know that the DN and CN pricing isn't bad at all.
Meaning even at $400-$700 for client CN and $1500-$2500 for DN is certainly cheaper.
There is nothing like this in the wireless market yet, period.
Fiber is the closest thing to it, and is probably required for most deployments at the edge of Terragraph.
Now, if you are talking retail fiber build, that is around $3000.00 per client all in to build.
I build for less than half that with my "at cost" model in my company doing Fiber.
But at amortized cost of less than $1000 per client gained, with 1Gbps FDX deliverable at the client (or more next year with bonding), this is a game changer.
I'm doing this as best I can at about 3-6 month ROI all in with 802.11AD equipment and it really isn't working well at scale.
If I can get this deployed at less than 9 month ROI then I'm all in.
ROI on Fiber is typically 30 years on bond. I've got that down to maybe 1/3 that at 10 years.
With my available infrastructure already in place to deliver 10Gbps fiber to the DN on one or more nodes in the mesh, at 9 month ROI I'm doing way better than fiber on costs.
And I'm able to deliver almost the same experience across nodes/mesh.
I'm certainly a LOT more dense and stable than anything else on the market at these speeds when we're talking wireless.
Where this doesn't work, is at distance. That's still the domain of the typical "WISP" last mile at 5GHz PtMP.
But we're really talking a different market entirely between Terragraph and 5GHz wireless.
According to the ads, the intention is to be able to deploy and replace fiber much faster and cheaper in densely populated areas. To achieve this in urban areas you need to put the devices really close to each other the same way as the high speed 5G architecture. Considering the amount of units needed to cover in a relatively small area, it will be a pretty heavy investment. You also need cover the cost to power all the mesh devices and the customers needs affordable CPEs. I really don't get the business modell. Anyone?
Prices according to Cambium: