I worked for 10 years as a researcher in quantum computation. Looking back, i would say that i see a mixed bag. On the negative side i have to say that many groups try to jump on whichever direction the most recent five papers in the field had been in, very often with little or no result at all. (if the Nature paper is out, the other group already followed the new path for five years).
On the positive side, we come to the other groups/leaders, which follow a direction which adresses aa problem until it's solved. In the superconducting QC field that would be for example (There are many other good and creative groups in the field) the group of John Martinis. They adressed the problems they saw over years in hard work (and that started in 2002 or earlier), at least such effort is usually rewarded in science on the long term.
But again on the negative side: the papers they managed to put in Nature or Science were focused on the final results of the engineering - the papers which really adressed the problem puzzeling the community for years, where they really found out how to reach the goal were published in Physical Review B, Physical Review Letters and some other Journals. (Phys. Rev. Lett. 93, 077003, Phys. Rev. Lett. 95, 210503, Phys. Rev. B 68, 224518, Phys. Rev. B 67, 094510, Phys. Rev. Lett. 89, 117901, Phys. Rev. B 77, 180508). The fact that enhancing the building blocks for a final result gives you much less impact factor than obtaining the final result make the stategy not be creative and hope for others to fix problems a reasonable one. Even catching a Nature paper every few years is enough for a conservative, non-abitious group leader, so you can burn a few postdocs in average, which you put up to the current topic, and if you a lucky, your results look accidentally good every few years, even if you did not contribute much to science.
