The "last mile" argument is true, your numbers don't disprove it because they don't represent a percentage of consumers with said broadband options.

The report's title is: Percentages of Households Located in Census Tracts Where Providers Report Residential Fixed-Location Connections of Various Speeds as of December 31, 2012

If the census tract where a household is located has access to 2+ broadband providers it satisfies this test, even if the household itself has access to less than 2. An entire tract that includes 2 cable companies that offer 10+ mbps but only offer exclusive service to locations would still satisfy this test even though 0 of the residences have access to 2+ services.

Yes it seems possible that they could be overestimating, but it is the most thorough report I have seen. And if their sample set is large enough, they should have accounted for it. Do you have any actual data that shows different numbers?

No I don't think they are overestimating, the graph just doesn't say what you said it does. It does not count the residences that have access to 2+ options, it only counts the number of houses that lie within census tracts that have 2+ broadband options available anywhere in said tract.

This means that even a residence with literally 0 options would be counted as a "2+ broadband options available" house if there were 2+ broadband options available anywhere in the census tract it lies within.

The numbers don't have to be refuted, they simply aren't numbers that indicate whether there is a problem or not because they answer a different. But unfortunately no, I don't have any data other than this report. I was super disappointed when it came out because I don't know of anything else.

You are right, what the report says is that 97% of people live in census tract where there are two broadband providers. What percent have access to the two is unknown, but absent any evidence to the contrary, and given the relatively small size of census tracts, it is reasonable to assume a majority have access to the providers in their tract. I'm open minded here, show me some data that shows a large group of people don't have access to at least 2 broadband ISPs.

Apartment buildings often make exclusive deals with cable companies for a variety of reasons. And even when they don't, it can be very hard to deploy new infrastructure anyway.

Yes of course, like I've said, the number is overstated, but by how much? How many people in the country live in apartments like you describe? 2%? 5%? 10%? It still doesn't make much of a dent into 97%

In most places it's still a duopoly, the choice between DSL from the single telephone company or cable internet from the single cable provider. It's just as bad as a monopoly.

Duopolies are not just as bad as a monopoly. Duopolies might not create perfect competition, but it normally produces pretty good competition.

AMD/Intel, Visa/Mastercard, Airbus/Boeing, etc. all have stiff competition.

And your local telecom and cable company are engaged in pretty stiff competition. At least Verizon and Comcast are where I'm at.