Your argument doesn't factor in that people need to eat, and they well, are making food.
As someone living in California I certainly appreciate the huge variety of great produce that's available because of CA agriculture, but even the almond farmers know this isn't going to last forever. They're all in a race to drill deeper wells than their neighbor and make a quick buck before the aquifer runs out. At that point we'll have no reserve at all and the party will grind to a halt.
The better idea would be to grow different crops here, not to stop agriculture altogether. That requires some kind of rationing.
The average US household consumes 11,000 KW-hours of energy per year. Let's say an average person uses 80-100 gallons of water a day, so a low end of about 30,000 gallons of water per year. One article [1] suggests desalination takes about 15kJ/kg water at the most efficient plants. That translates to 473 kilowatt hours of energy required per household each year:
30000 gallons * 3.79 kg/gallon * 15 kilojoules
So that's about a 4% increase in household energy usage just for domestic water for one person (domestic is less than 10% of our water use in CA). I'm not sure what conclusion to draw from this, but I suspect that it isn't practical to use desalination as a primary source of water. I also expect that much salt would become an environmental problem to dispose of. Once we drain our aquifer, the water shortage will be really severe and we'll need to cut back on agriculture regardless of desalination, but it could help a lot in urban areas.
[1]: http://www.desware.net/Energy-Requirements-Desalination-Proc....
Percentage of household use is a strange measure to use here - Ivanpah plant produces nearly 400 MW and it’s relatively small plant. We have a lot of desert to go around.
Producing more solar for the regular grid is not very practical because we can’t use all that up during daylight. If you have dedicated use for desalination that is concurrent with peak production of power, you can bring online a lot more production without risking destabilizing your grid. In effect, desalination can be used to balance the excess supply of solar - the more sunlight we get the more water we can produce. Water is relatively easy to store, modulating the demand to match the available excess supply.
Almonds are on the high side compared to other plants but their water usage per calorie is much less than animal ag.
Also consider that almonds flourish in the mediterranean type of climate that California has, whereas dairy and beef (and their feed crops) can be raised almost anywhere.
The world produces far more calories of food than is needed, so water impact per calorie is a dumb metric since calories aren't the critical figure of merit.
What matters is water impact per $ of dependent economic activity.
You're missing the point that we are talking about regulating water usage external to the industry (in which people should and do care about water usage as it relates to sustainability of California as a whole), and water usage internal to the industry (in which an individual farmer cares about yields, efficiency, and economic value of crops).
Internal to farming the entire reason we have so many almonds in CA is exactly because it's a high-efficiency and high-value crop.
https://slate.com/business/2015/04/almonds-in-california-the...
"In a real sense, the almond industry is the future of California agriculture—high value, high efficiency, but still high consuming. That’s not necessarily a bad thing, as it makes most sense to use a scarce resource for the highest value application possible."
Well, no, because I was positing a better metric for competing agricultural uses, not the one-true-metric for all uses and, also, because even if it was the one true metric for all uses, unless water was the limiting factor for urban economic production, while it kight imply all cuts needed fron current consumption should cone from ag, it wouldn't imply additional cuts to ag to increase urban alotment. And, while the reverse allocation of cuts would cut all urban use in a foreseeable shortfall (because its such a small share to start with) any shortfall that would require ebough cuts to equal all agricultural use would be a a cataclysmic disaster no matte how the cuts were apportioned (and most would cone from agriculture because, even after you stopped giving any to people—whom you need for agriculture, anyway—you’d barely have dented the necessary total cuts.