A project like this will have many considerations that don't fit into a Hacker News comment. I'd be surprised if a full analysis was less than 200 pages of detailed hard work, including performing actual research to answer questions such as "what exactly is the impact of a surface magnetic field of 0.34 teslas on the local water flow?" and "is marine life impacted by the local magnetic field exceeding Earth's natural field within about 10 km?"
I would be surprised if the very slow shift to Earth's poles has any impact, but yes, that too, even if only to rule it out.
Temperature, however, isn't one of those questions. Deep water has a near-constant temperature, and even if it didn't, solving for this is already well-studied for continuously welded railways.
Likewise, because it's underwater, we can make them in neutrally buoyant sections. Could even make them as two parallel and closely-spaced tracks, so it can built as an underwater railway that delivers its own power and transport during its own construction.
> due to the very real possibilty of another technology bieng adopted and implimented first.
Not hugely likely, IMO.
Long term I can see an argument for doing this as an orbital ring — basically the same idea but make it about π*200km longer so it can be 100km above ground everywhere, spin it so fast it's slightly above orbital speed, magnetically levitate hoops around it that are tethered to the ground and transmit power.
Buuuuut orbital rings are only at technology readiness level 2, so nowhere near ready to consider:
https://en.wikipedia.org/wiki/Non-rocket_spacelaunch
https://en.wikipedia.org/wiki/Technology_readiness_level
(For example, an orbital ring has to be build by thickening rather than lengthening to be affordable: it was conceived as a non-rocket launch mechanism, and bootstrapping it by having a thin version lift the material for making it thicker and stronger was a big part of the pricing when it was originally conceived).
Whereas a big fat metal wire is at the level of "do sufficient geological and biological surveys to make sure there's an actual route, and accept that the answer may in fact be 'no'" — we *definitely* know how to build a big fat wire (or many smaller wires). We just don't know if we *should*.