Sorry, I guess I was not clear. I didn't mean what would be the justification for switching from F to C now in the US.
I meant what was the justification for adopting C initially for the metric system instead of using F?
For other metric units, there were serious problems with the pre-metric units. For example, Wikipedia says that by the time of the French Revolution, the existing system had become impractical for trade.
It doesn't say why it had become impractical, but my guess would be it had a lot to do with no good standard values for the units. If the units are not well defined in a way that gives the same values everywhere, it makes long distance trade harder.
I base this guess on doing some research once to figure out why they did not defined the meter to make conversion to Imperial easier. Since they were free to define the meter any way they wanted, they could have made 1 meter == 1 yard, which would have made conversions of long distances easier. Or they could have defined the millimeter as 1/25th of an inch, which would have made conversions of short distances easier than the 1/25.4th of an inch that we ended up with.
As far as I was able to determine, this was never an option because, as far as I could find, there was no widespread agreed definition of inches, feet, yards, etc. All that was standard was the relationships among them (1 foot == 12 inches, 3 feet == 1 yard, and so on).
There probably was no way to fix that without adopting a completely new unit, because if they tried to just make a new standard for, say, the yard I'm sure it would have gotten bogged down in arguing over whose yard to base it on. The French would think everyone should adopt their yard, the English would think everyone should adopt theirs, and so on.
So, a new unit, based on something not tied to any one region or nation, makes sense.
Temperature did not suffer from this, or did not suffer from it in a way that was not easily fixable. F was based on a reproducible physical 0 point based on the temperature of a brine with a self-stabilizing temperature. The only problem F had was that there had been some waffling over the other calibration point, which had been fixed well before the first attempts at a metric system.
So...why did they pick C?