1) Bury the lines. Basically isothermal at depth.
2) insulate the tube and maintain stable temperature
3) use materials that don’t drastically change length with temperature
3) allow the tubes to change length. Put them on rollers.
4) use vacuum expansion joints just like us common in vacuum engineering.
People act like vacuum is this new thing. Physicists (well, Engineers/technicians hired by physics projects) have been making really long tubes with extremely high vacuum and incredibly high tolerances. This is easy in comparison.
Burying the lines would already make it an order of magnitude more expensive than what was promised. And now you have to compensate for ground movements, since your tube is fat and hundreds of kilometers in length, which makes it even more expensive.
Insulating the tubes to maintain temperature would be incredibly costly, and if there ever was a failure or natural disaster that prevented you from keeping up the power need, your tube will break.
Materials that have such little temperature change coefficients are much too expensive to build a 600km long multiple meter wide tube with.
If you allow the tubes to change length, your stations will move, sometimes in unpredictable directions. You also risk gigantic stresses caused by temperature differentials causing bending forces.
Vacuum expansion joints of that scale, for such a length, with such reliability are literally unheard of. And would likely be insanely expensive.
2) Kaa-ching! Both upfront and ongoing.
3) Kaa-ching! There is reason materials such as Invar are not found in railways, and cost is only one of them.
4) There is no precedent to using vacuum seals at that scale (or maintaining vacuum in that sort of volume).
Elon Musk just says "let it be suspended on roller supports so the terminals slide dozens of meters". Nothing wrong with that surely.