I often fantasize that when I will get my work to successful end, then I will have new possibilities open – and that maybe I will even receive scientific job offers. The work is obtaining new important results in computational sciences.
But then I get to the ground, think that giving someone job is rather a necessity for the employer – and not an act of appreciation or even not an act of support. Then, my results are important and useful, so someone may need them. However, the results can be understood, used and further developed by some other scientist. So it will not be necessarily me, who will get the job.
One related example that comes to my mind is Stephen Wolfram, who is independent because he is earning money himself. So he was not appreciated by someone, instead he won his share in software market.
Sadly, there is not a strict correlation between doing good science and being financially successful. It is entirely possible to do "creative, important work" and still not be well rewarded for it. For instance, one could have published these results in an obscure journal that very few people will read. Or, as another example, the researcher might be a poor "salesman," unable to convince readers and fellow scientists of the merit of his work.
In general, you need a combination of both networking and technical skills to forge a successful career as a researcher: the contacts will help you get interviews for jobs; your technical knowledge will get you the job.