Alternately, maybe there's something we don't see in the one-line project description from an internet comment.
edit: A quick googling answered that - they're buying the power for a data center in Nebraska, not in California.
[0]: https://en.m.wikipedia.org/wiki/High-voltage_direct_current#...
That said, I agree that 5% sounds close for the transmission losses alone (excluding distribution), and 5% could be easily made up, like you said, by many things, like being next to a wind turbine factory/cheaper land/cheaper labor/tax breaks/many things.
Interesting note: I found here [0] that silver would be a feasible conductor for HVDC lines if prices were lower. On an example line, it would save $19M/yr but cost $18.5B to build, so 1000 year payback time, which is unrealistic. But! If the price of aluminum increased substantially (for example: say we discover smelting releases mega-carcinogens and it's banned) and the price of silver fell (silver is now 4x as expensive as it was in 2003, so I feel like it could fall fairly precipitously) then we could in theory have silver power lines. One thing to remember, though, is that it's only the "skin" of the conductor made of copper/aluminum/silver - the skin carries 90% of the power, but the bulk of the conductor's mass is the cheap steel core that doesn't need to be as conductive because it only carries 10% of the power.
Rather than spending $3B to build out wind or other renewable power locally.
First, hydro power is very stable, and Quebec has a surplus, so they could probably get a good price deal for an ongoing relationship. The construction cost is nothing, compared to the ongoing cost. What does that Canadian power cost, vs the cost of wind? They could be looking at tremendous savings over the next few decades.
Second, this appears to be a New York City decision, not a New York state decision. Does NYC have enough viable land to build a gigawatt (or more like 3-4gw, considering load variation) of wind? Probably not.