> progress will likely only accelerate

I'm guessing this is part of the whole idea; to get data before its actually doable, and the cost is worth the early data to help with future development