Actually, I think No.s 2 and 3.
There is no climb forcast page, so I Think the FMC assumes linear variation between current wind and wind at altitude. As this is not the case with a jetstream, climbing as quickley as possible into the tailwind, or getting as many track-miles as possible before climbing into the tailwind, would seem the most efficient way to fly.