The thing is, we already paid that bill years ago. Taxpayers gave something like $200 Billion to the Baby Bell telcos in exchange for the promise of building a broadband network. We paid, then they didn't build it.
The number is computed by comparing Baby Bell profit margins after deregulation to "utilities" as a whole and calling anything above the average for "utilities" as "excess profits" that were a "taxpayer subsidy" to the Baby Bells. It is based on two fundamentally incorrect ideas:
1) That the expected rates of return of companies in an industry with exploding demand, like telecommunications was in the 1990s, should be comparable to the rates of return for your water or power company. But water and power utilities use 50-100 year old infrastructure to cater to extremely stable demand. That's exactly the opposite of what was happening in the telecom sector in the late 1990s.
2) That regulating telecom rates so those companies were no more profitable than water or power companies would not have affected investment. There is a reason our water system gets a "D" from the American Society of Civil Engineers: http://www.infrastructurereportcard.org/water-infrastructure.
Imagine comparing the profit margins of smartphone manufacturers in 2007-2017 against the average profits of "phone manufacturers" before that time, and calling everything over that average a "taxpayer subsidy." Alternatively, imagine telling phone companies that they could not make any profits beyond the average profitability of the phone industry from 1995-2005. Do you think we'd have iPhones today?
I'd be surprised if the real number was even a net subsidy rather than a net tax. We impose all sorts of extra taxes and costs on the telecommunications that we do not impose on other industries. For example, state and local cable TV franchise fees alone (typically 5% of gross revenue) amount to $5-6 billion per year in the U.S. In comparison, the American Recovery and Reinvestment Act had $7 billion in telecommunications subsidies, and that was a one-time shot.
Then there is the economic deadweight loss created by implicit and explicit cross-subsidies. There is a 17-18% tax on telecommunications services.[1] That is not a net tax, because the money is redistributed within the industry, but it increases prices, which decreases demand, which causes a net deadweight loss. Then there is implicit cross-subsidies in the form of build-out requirements. In nearly every city, you are not permitted to only build out in the neighborhoods where you can count on having enough subscribers to recoup the cost of that part of the build. You have to build out everywhere. The end result is that prices go up in neighborhoods with more potential subscribers in order to subsidize the deployment to neighborhoods with fewer potential subscribers. Again, making the product more expensive decreases demand and imposes a deadweight loss.
The ideas behind these policies are mostly well-intentioned. We want as a society to provide connectivity to rural areas and poor neighborhoods. But the implementations are driven by politics rather than any attempt to achieve the desired end result in the most economically efficient way. We don't pay for SNAP benefits with a tax on supermarkets. We impose a general tax on everybody, to avoid distorting the economy in favor or against any particular industry.
[1] That tax pays for almost all the "subsidies" you hear telecom companies getting. It's like imposing an 17-18% tax on all smartphones, and giving the money to companies building cheap smart phones for poor people. Nobody would call that a "subsidy" to the smart phone industry (it's a cross-subsidy within the industry).
There are cheaper optical water level sensors in the aquarium trade that might work for you. Check out the Tunze Osmolator as an example. It could easily be connected to a GPIO on the pi.
That's a very interesting idea. Thanks for the suggestion. It is still invasive since I have to cut the intake pipe and insert a section with the sensor installed in it. But it is more flexible since I can install it in any size pipe. And for my use case of a 1 inch input pipe it can be cheaper than my current solution. For example https://www.aliexpress.com/item/15mA-5V-Optical-Infrared-Wat...
Although, I wonder if scaling (as in hard water scale) is still going to be an issue.
https://www.techdirt.com/articles/20060131/2021240.shtml