Your comment presents an excessively gloomy view of academia. I am yet to find a productive graduate student / post doc in engineering or CS at a decent univers

Your comment presents an excessively gloomy view of academia. I am yet to find a productive graduate student / post doc in engineering or CS at a decent university, who hasn't been able to find a tenure track position. I think academic jobs have the advantage of almost full freedom to do what you want. In addition, there are opportunities to exercise entrepreneurial skills via grant-writing. Besides, for engineering / CS PhDs, it is fairly easy to transition out of academia since there are more than enough job openings. I think the main problem with tenure track positions is that the bar is often too high compared to the industry, and you require more 'stars aligning' than with industry jobs- your narrow field must be hot the year you are applying, you must have your strongest resume within the December / January deadlines that most job postings have, etc. Industry R&D also has its pitfalls. The primary metric in most industry R&D labs (Microsoft, Mitsubishi ERL, etc. are the known exceptions) is $s, and there are more stakeholders in getting $s (customers, managerial caprice, etc.) than in publishing a paper or 'winning' a grant (which are usually the metrics in academia).