HSR would be easier with a land value tax - there would be a negative incentive to hoard land, so land acquisition costs would be minimised.
California is a great example: HSR battles and Prop 13 revenue dilemmas are things that land value taxes are directly intended to resolve in favor of the public wealth. But both of these phenomena have strong public support propping them up, directly and indirectly. Prop 13 was passed precisely because the public was concerned about the threat of people being priced out of their homes, and that sentiment is at least as strong today as it was back then. Land value taxes would systematize homeowners' worst fears, and they won't be placated by arguments or even evidence that shows that they're better off on average.
I'm tempted to say that people are just irrational or that the public often wants contradictory things, but that would be lazy--true but lazy. In any event, my point is that resolving these things requires more than simply engineering a system that works better on paper. Incentives have to be aligned not just in dollars, but also in votes; not just at the outset, but also on an ongoing basis. Georgist land value taxation schemes are fundamentally predicated on consistent, long-term application. Once exceptions are made it crumbles because not only do perceptions of unfairness increase, the utility of the regime declines. This is true of alot of things, even traditional land taxation, but it's especially true of land value taxes.
[1] Or at least I'll concede that point. It's not obvious to me if and how prospective public works projects would figure into land values under a Georgist system, especially in the rural areas where HSR battles seem to be the worst. (AFAIK, the more urban areas in and around Fresno were acquired and bulldozed rather quickly.) Particularly with land value taxation, the devil is in the details.