Clearly not true. We can design individual tools to do individual jobs, but we can't design and deploy a general purpose scientific instrument that can react to new facts on the scene. We have to design and send an entirely new device in most cases. And forget it if something goes wrong. I somehow think an astronaut could have yanked on the "umbrella" antenna of the Galileo if they'd been along for the ride.
More importantly just because isn't a rational position and worthy of debate. We need to get out there in order to establish a presence other than on earth for our species. We CAN'T stay here much longer, relatively speaking. We'll either kill off the ecosystem or it will kill us, or some external force will kill us both off. We have to go on to the next environment and make it ours, or we need to curl up and die. I prefer to go on. Now, do we need to do that Tuesday? Only to satisfy my own impatience, and to make sure that the folks that go speak english and not mandarin or hindi.
Assumes the failure rate for manned flight is zero. That is not the case as we know. Each Shuttle launch costs about $400 million. The Mars Pathfinder costs about $260 million. We are paid the Russians $473 million to use MIR. Apollo cost $117 billion. Voyager, Galileo and Pathfinder combined were just over $2 billion. Much more bang for the buck with unmanned vehicles.
We use robots for deep sea exploration and commerical activity. Why not the same for space? At least until we have the technology to do more than what we can now.