1) How much infrastructure do you need to support manned versus robotic exploration of mars?
OK, for manned presence you have to have a safety margin. This safety margin requires massive redundancy. That means you not only have to have the basics of food, air, radiation shielding, gravity remediation equipment, and transport capability, but failsafe redundant mechanisms to insure survival where there is no possibility of timely rescue missions or resupply. Everything has to not only be designed for human survival (already massively costly) but multiply redundant and multiply costly. You say lets have a habitat that can sustain itself for years( we have been unable to do that in low earth orbit!, not even close).

Lets contrast that with robotic exploration. Is the loss of a robot costly? Yes, but it in no way compares to the cost to space exploration to the loss of an astronaut. Also, there is no moral requirement to insure survival of robots except in a cold calculating tradeoff of maximizing effective science return. We've already had rovers on mars for YEARS with minimal cost. These are only 2nd generation rovers! With each generation (every 2 years) it is probably that we can increase science return several fold (conservatively).

3) Realistically, what is necessary for a true SUSTAINED AFFORDABLE manned presence on mars?
To really have a sustained presence how many missions are required? What would be the size of the sustained "research station"? Obviously we would have to "live off the land" greatly, because the cost of transport of support survival materials will be almost unbelievably expensive. How much is yet to be done to maximize living off the land on Mars. I haven't seen any plans that really address the details (we don't even know all the questions to ask) except in science fiction books. There have been defined some areas where we know we have to find more knowledge, before we can even think about sustained presence on Mars.

4) What are all the steps necessary before we can safely launch a manned mission?
You have to understand dust, radiation, gravity, psychological isolation. You have to design a unique mission never done before, with failsafe redundant capabilities. You have to understand, a tremendous amount about Mars (resources, accessibility to same). You will have to have failsafe power supplies and a large amount of robotic support for transport and exploration. The potential sites will be very limited for both safe landing that provide resources. Will those same sites be the best sites for exploration? How do we get all this information.

5) What can human presence really add to understanding Mars?
Do you really think a handful of human beings are going to do the research to understand Mars on Mars? This is absurd. The brain resources and huge science machines to truly understand Mars will remain on Earth for the forseeable future. Sample returns will be required whether you use robots or humans. Sample returns per sample return will be vastly cheaper done robotically versus with humans.

How much would humans on mars use teleoperation rather than personal presence at exploration sites? The overhead required for manned excursions even once on mars, will far exceed its usefulness. Men on mars will just end up using robots to do their work for them anyway. Could this be done off Mars? At the very least could it be done from phobos or in orbit?

Let me play out a scenario for using a 4th or 5th generation rover teleoperated from Earth. A human being enters a full 3D enhanced virtual environment. He works through a simulated environment moving the robotic probe from place to place performing a chore list. This is a true simulation so that when a rock is removed what is underneath can not be shown in real time. However, directions to move a rock, dig under it and sample can easily be entered. Results from the chores are updated as they occur, whether it updated visuals returned from Mars or processed sampling results. Continuous feedback comes from the rover as it receives and excecutes commands. Using multi-spectral surveys already updated, some interesting rocks are spotted. The human operator approaches each rock and does a first pass close up inspection that only takes moments (i.e. brushed, and visually magnified to see grains, etc, and any other scans that don't require lingering. He does this with each of the rocks at the site, then he tops a rise overlooking the site and performs a multispectral survey of an adjacent site. By the time this is finished results from his preliminary rock studies are arriving, and one of the rocks demands sampling or further spectrometric analysis. He returns to do this. The simulation is slowed whenever the rover is slower than the simulation required. Occasionally when there is a problem the simulation will be halted and there will be an inefficient interval that requires some multiple 2 way communication, but these should be rare events as the rovers are very capable of following general commands and executing them with few problems.

When the human finishes a shift, another takes over. The rover is operated continuously (even at night), because it is nuclear powered or has enough battery power to sustain it. This is repeated through dozens of rovers spread over the planet, some with unique characteristics (to scramble down canyon walls) or to go down old volcanic vents. The rovers requre no sleep or food, as long as sufficient thermal protection and power courses through them there is little reason to worry. They do not need to return to a base and can remain in the field potentially far away from their initial landing spot. Your point that a human could do this exploration faster is based on what? So far the robot has done it much faster. Even on mars, the main thing the human would have over a one of our current rovers is that he could walk faster from rock to rock. Will that be able to be said in 5 to 10 years? I seriously doubt it. In fact I will bet you that a robot will be faster than a human. This is like Paul Bunyon fighting trains and chainsaws with his axe. It cannot abide.

Samples are gathered from various points on the planet and combined at a few key rendevous points. From these points they are returned to mars orbit to once again rendeveous with the sample return ship. These ships have no requirements to provide food or oxygen, space or radiation shielding so they can be relatively low mass and can maximize the size of the sample return.

6) How much do we need to understand mars, before we can do 3?
What about other things needed other than sample return, like deep drilling, radar sounding, geophysical exploration seismically all can be done through teleoperation more cost effectively than with in situ humans.

7) What about one of the main reasons for going to Mars, biological sampling?
Even with biological sampling, wouldn't it be better to transportant 10,000 pounds of science laboratory to do remote study of potentially biological samples than a human with a microscope. What is a human going to see with his naked eye when it comes to biological samples on Mars? Plus you avoid contamination issues with biological sample return.
------------------------

OK, lets summarize. Exploration of Mars is vastly cheaper if done robotically (teleoperated), rather than with humans. We are talking probably 2 orders of magnitude, so instead of 10-30 billion we are talking over a TRILLION. Really the only reason to have humans go to mars if is you could establish a real colony. In order to establish a real colony you really need to explore mars FIRST at a level that already requires the ROBOTIC exploration. So, why can't we bite this bullet and truly fund robotic exploration at levels more indicative of its importance in its place in the critical path to manned presence?

Funding for manned mission to mars should be relegated to longterm studies of radiation, gravity, psychological and other issues that we have a dearth of knowledge about, that must be more fully understood before we can even get started.

-------------------------
A lot of this same stuff applies to the upcoming Moondoggle manned return to the moon, which is 95% political and 5% practical. Telepresence on the moon would be relatively easy as it is only light seconds away from earth. Once again, robotic exploration should be emphasized before manned landings. If we can't build everything we need to survive and explore without men actually there it is not designed right. Once a base is built and the robotic presence established, then you can populate the base if it is really necessary (the only reason I can come up with is for political reasons).
-------------------------
Finally, if we really want to get a manned presence off world then we need to lower launch costs, lower launch costs, and lower launch costs. Second we need to get space to pay for itself (the space tourism industry is probably the best option for this currently). Once a feedback loop is established, bootstrapping our way to sustainable presence can happen. A top down NASA super mission attempting the same is doomed to cost overruns and failure.