I imagine public Internet or Starlink would suffice to upload daily readings. But sure, that's speculation from someone who knows nothing about how it works.
The costs of this equipment get amortized for the duration of the equipment's useful life, so you can't just throw that away. Even if we do though, $10 is still too low of an estimate. For sake of argument, let's assume there was one guy running the entire network, part time, for half an hour per day. He will at minimum be making $30 an hour, add another %25 on top of that to account for overhead costs, so $37.50. $37.50 / 2 = $18.75. We're already over your threshold.
If other estimates that this equipment is in the realm of $200,000 per node to set up (that's presumably not just the cost of the server and other equipment but also the cost of delivering the equipment and standing it all up in the embassy), assuming 100 nodes you're talking about a $200,000,000 program just in capital costs. Amortized over a 10 year life span, assuming linear depreciation, the amortization costs alone are $3,650 per day, and that isn't including maintenance.
And yes, $10 is hyperbole, but we have AWOS (automatic weather observation systems) at airports all over the USA. I wonder how much it costs to transmit data from those.
If not, why isn't tailscale free for everyone?