This was the biggest lesson I learnt when doing my PhD. One of the sections I had to look at what factors led to successful software projects within companies. This could be anything, from going with Open Source, buying the right software from the right vendor, management involvement, 10X developer in the company or even something as simple as user training. You can practically pick whatever factor suits your bias and you will find the data that supports your argument.
You are right. The problem comes when it is time to weight which one is most important when making a decision. Open source Vs of the shelf all these "true"facts are brought to the table. It often ends up being what head honcho wants or going with IBM/Microsoft. No one ever got fired for making that decision
