I don't buy it. You're comparing the energy inputs required to support large-scale tissue engineering operations with those needed for grazing cattle on pasture. It's not even close.
If it was just pastures I would agree. But 30-40% of all corn produced in the US (the largest corn producer in the world) is used to feed lifestock [1]. That's a lot of perfectly good farmland that could directly produce food for human consumption.
Even if the lab process takes more energy, as long as producing the growth medium takes up less farmland that is a win towards the goal of having enough farmland to feed the world.