A constantly asked question about Free Software is how big the commercial influences on Free Software are. Empirical studies about this topic more often than not contradict each other. It seems clear that it is methodologically difficult to set up a good survey on this issue in the first place.
Part of these difficulties probably come from the fact that peer production and capitalist production are closely interwoven and so it is very hard to make useful separations at all.
On the other hand peer production influences the companies as well. It would be thus also interesting to study these influences as well - though this is probably even harder.
It would be great if there would be an empirical study on Free Software products which answers the questions most relevant to Oekonux. May be some researcher out there finds this and does the world the favor.
Here are some requirements which come to my mind:
Consider small projects also
Most of the studies which actively approach Free Software projects do so by looking at the big and huge projects - like the Linux kernel, Eclipse, OpenOffice.org, Mozilla, and so on. However, a good part of for instance a complete Linux based operating system are small programs such as grep or Perl modules like in the CPAN. These small programs are often maintained by a few people if not a single person.
A useful survey needs to make sure that those projects are considered as much as the big ones are.
Careful consideration of paid time
What people confuses mostly is that there are Free Software developers which are paid by their companies for their development efforts. Or to be more exact: That they work for Free Software during their paid time. However, that people do something during their paid time does not say everything about the relation of their employers to what they do - think of coffee breaks for an example.
I think there are a number of types of Free Software development during paid time which needs to be distinguished:
That is when people work on Free Software on their spare time without knowledge of the management. This can be the case when you use a certain Free tool during your job and improve it as a side effect to your normal work and give the improvements back to the community. This also applies to Free tools developed during a job.
There may be times when people are not working on a job project because there is no order at the moment. In these times management may give people the opportunity to work at their own projects which may be Free Software. Google even has an official 20% share of such time.
If you work in the software sector then you know that often there are tools you use but which are not of competetive interest to your customer. For instance in a big company you may develop a Wiki engine for the Intranet but the core business of the company is in a totally different sector - such as finances for instance. In that case the management may allow that in-house development be public Free Software.
This is the case when a company officially engages in a Free Software project because it has manifest interests in this project. For instance if a processor vendor engages in the Linux kernel to make it run on his processors. Then the employees not only officially work for the company in the Free Software project but also this is relevant to competition.
Only in the last case I see the danger of alienated influences from paid labor. And even then it should be distinguished whether these influences are more useful or more harmful for the community.
Careful consideration of non-paid time
If develop Free Software during non-paid time then it seems to be clear that this must be based on Selbstentfaltung. However, there might be cases where people during their non-paid time they do things in the interest of commercial interests - be it in their own interest or in the interest of their company.
A useful survey must try to make a distinction here as well as it does for paid time.
Different commercial interests
It would also be useful to distinguish between the source of commercial interests. Is it because a developer wants to learn something or her employer wants her to do so? Or is it because the developed Free Software should fulfill a mission critical commercial interest.
This is list is far from complete. Feel free to add the points relevant to you.