I'm afraid I give the nod to robots too.

Not neccessarily with today's technology, but imagining what improvements we'll see over the next 20 to 30 years.

Right now, the two biggest limiting factors in robotic missions (taking Mars surface exploration as the example) are insufficient autonomy and power constraints.

The two biggest limiting factors for human missions are the cost of moving large masses between planets, and of designing long-term closed life support systems.

For robots, computing ability is getting better FAST. If the rovers were able to do better autonomous route selection, and had more automation on using the IDD ("arm"), they'd get a lot more accomplished per sol. The ability to do this is improving rapidly as hardware improves, and we gain more software experience. Plus, robots that move autonomously are becoming more and more common for Earth-based commercial and military applications, meaning there are going to be a lot more engineers developing such systems.

Power will be a problem so long as we rely on solar cells, but if O'Keefe can pull off the politics of using nukes again, then rovers powered by RTGs or the like will have power to move at decent speeds.

So by 2024, you probably could have a rover of MER's size and mass accomplish the same tasks in a week or two, not 45 days. And even humans would still have to wait hours for certain measurements; APXS data collection takes hours, and would if a human set it up, too.

In contrast, propulsion and life support are advancing much more slowly.

Whatever you think about the advantage of robots vs humans today, the balance will be steadily improving in the robot's favor.