And what exactly could a human geologist do that a rover (built with current technology), coupled with a competent geology team on Earth couldn't?
So you don't actually want to think about it? The answers are the same as here on Earth. Sometimes a robot is necessary but we don't use them when we don't have to because working through them is incredibly awkward. There is no geologist that wants to work remotely when he can work onsite. Plus there is a LOT more to do than simply geology on Mars.
1) Speed - a human can work faster on site than humans working through remotely operated devices. When latency averages 13 minutes each way humans on site are a LOT faster. A human on site could accomplish vastly more in a shorter amount of time.
2) I defy you to find me an end effector for any robot that is as useful as a human hand attached to a real live human.
3) Notice and investigate things the robot wasn't designed to address.
4) Repair equipment that breaks
5) Utilize local resources in ways robots cannot
6) Do things other than geology
7) Design and implement tools on site
So a $100bn Mars mission is going to deliver $10,000bn in R&D payback that couldn't have been done without sending humans to Mars?
I said that the R&D payback would be much larger for a manned mission than for a robotic mission (or even a large number of robotic missions) which is true. I didn't specify any dollar amounts - The 100X number is just made up to get the point across though it seems to have whooshed by you. A LOT more technology would have to be developed for a manned mission and as a result there would be a much larger R&D payback. For a manned mission we would need all the robotics research PLUS life support, radiation shielding, food supply, medical technology, remote manufacturing, mining and much more. It's a far bigger, more complicated mission with a lot more R&D requirements. Bigger R&D will result in bigger economic payback.