I remember in Transport Tycoon, under certain conditions, you were no longer allowed to bulldoze willy nilly. If you wanted to put in a new bus station and were out of favour with the populace, you could try the old fashioned way: with bribes (and tree planting).
Similar situation - guy remembers a catchy song, lyrics and all, but can’t find any trace of the band or the music. I won’t say more than that as it’s worth listening to them tell the story (The Guardian article also doesn’t spoil it).
Clear skies happen regularly enough. You can see refractive phenomena frequently (sun dogs, moon halos and the like). Same for night, plenty of clear skies and auroras during the winter when it’s dark enough to see them.
IceCube sends housekeeping and priority transient events (like a supernova) over satellite, selected high quality events get synced over satellite daily and everything else gets stored and shipped out in the summer. There is a redundancy everywhere - it’s a very thoughtfully designed system.
They didn't just travel on-site, they over-wintered, which means travelling to the site and then staying through the worst weather, while it's completely dark, without any practical means of leaving for months.
I feel obliged to say that while it gets cold over winter, the weather isn’t particularly bad by Antarctic standards; the coast is much worse. In places like McMurdo it can get bad enough that travel between buildings is prohibited, South Pole not so much. I also wintered for SPT which involves one of us going out to the telescope, daily. There were only a couple of days when it was marginal. Even then, winds and low visibility usually means cloud cover and warmer weather (-40 vs -90). The worst days are windy with clear skies.
As for darkness; consider a single day is stretched out over the year. There’s only a two months of astronomical night and the moon is up every couple weeks. You’re correct though, that the station is isolated between mid Feb and late October.
Neutrinos need a large detector volume for efficiency because they interact so rarely. You can’t detect them directly so you need a transparent medium to detect their collision byproducts. Good detector mediums are water and ice, and are underground to minimise background light. There are relatively few places you can do this. Mine caverns and under the sea are the most common, but marine detectors are notoriously hard to build. Francis Halzen proposed using ice. At the Pole, the glacial plateau is 2 miles high and the breakthrough was confirming that the ice is in fact highly transparent if you go deep enough.
Why the pole specifically? You could probably build a second IceCube 100 miles away, but how are you going to get that materials there? Pole has a skiway for large aircraft and infrastructure to house a large number of people. The traverse (SPoT) only became operational near the end of construction - the initial holes were drilled in 2005 and almost everything had to be flown in.
Telescope In The Ice does a great job of explaining the history and science behind the experiment.
The biggest challenge is finding ice without dust layers from historical volcanic activity, and the South Pole is particularly clean because of it's remoteness as well as the prevailing wind directions. There is, however, one big dust layer that causes signal distortion.