For a single developer, who is developing a .NET application, what is the difference in cost between:
[Note: Edited heavily. If I've changed the meaning, feel free to roll back. - Jon]
You could use the Express versions of VS.NET and SQL Server for free. You just need a license for the OS.
That depends on the knowledge and experience of the developer in regard to these tools. Plus, for Windows add the costs for every software you need to buy. Maintenance costs should be identical, at least they're not predictable.
In short: Use whatever you feel more familiar with.
Because its about .net: Of course, Microsoft's implementation is the original one, is it older and probably more stable. I don't think this will matter, however. Mono works really fluently.
It depends, mostly on people and on your specs.
In general, this decision involves trading money for time. OSS tools will save you money, but you will likely spend more time figuring out how to get everything to work smoothly together.
Basically, you should ask yourself "Do I have money, or do I have time?" and then choose accordingly.
This is more of a suplementary answer then a definite answer but I still think it has merits. The question has two answers:
The cost of development tools are roughly the same in a sense. If you go with closed-source you are paying for things upfront (like buying visual studio and free tech support.) If you go with open-source you will pay in the future (free development tools but you have to paid for great tech support [need for extremely hard problems or nasty bugs] and contributions to the original developers as a token of appreciation.) It is a matter of picking your poisons. The same apply if you mix-and-match the two.
The cost of development tools is irrelevant. The big cost of software is not the tools. You can get great tools for a single developer for like $2,000. However, a rockstar Software Engineer can demand a salary of $100,000 a year. Clearly, the cost of development is not in the tools but in the Software Engineers.
The main cost in software development tends to be people. There are exceptions, where the tools cost a year's salary per seat, but these are rare and specialized. (Part of this is of course due to F/OSS tools: there are decent development tools for most environments for free, so companies like Microsoft can't charge too much. Development software was a lot more expensive, in constant dollars, thirty years ago.)
Therefore, the important question is what set of tools will allow your people to work faster, and I don't know of studies comparing various web development tools. (My personal experience is that Microsoft tools are slicker than F/OS tools, but the slick stuff only goes so far, and after that I prefer F/OS in general. Your experience and preferences may of course vary.) If using one set of tools allows a team to finish a project a week faster, that set of tools has almost certainly paid for itself.
Deployment is another issue, which you don't mention. On some projects, deployment is expensive, and saving money there is important. On others, a couple of servers is all that's needed, and the cost of commercial software is negligible compared to the cost of initial development.
Preparing for future projects is yet another issue. A team that's going to do a lot of web projects will do best by spending not only money but time to get familiar with the best available tools (and I don't know enough to suggest what those are). An ad hoc team doing a one-off project should generally avoid retraining.
The best choice in the case you describe is probably what your developers are familiar with, Microsoft or LAMP or Mono or whatever, since they'll be able to get going faster and finish the project earlier, saving on the people costs that are going to dominate your project.