I would wager it's closer to reality than the number for rail. We've built enough highways that we have a pretty good idea what the costs are. High speed rail has fewer companies bidding on it and less experienced overseers.

My guess would be that the rail line would need entirely new right of way which would radically increase the variability of the cost and may result in it being down right impossible. (Speaking as someone who closely watched the drama around getting new right away along I5 for high voltage transmission) While the highway expansion probably already has the land they need for most of the project.

It's true for every major project, of any type. I can't think of any multi-year, >$10B software projects that ended up being completed on-time and on-budget, either.