If they all go and work for industry, it works to an extent Wow and it may be that more people need training in the subject; and better salaries to be offered to them for them to become teachers.
I don't like the work concentration though. The UK and its companies are finding it useful to invest in science and engineering. It is useful in that novel material and knowledge are produced that will give them a competitive edge. It is useful in that it trains people to go to industry with the required skills; it provides an answer to a demand. If you do a PhD in engineering in a university I know well and in a subject I know well and/or on applications I know well
, companies will line up to hire you because of those skills. They are also willing to fund PhDs. An example of such success, which hires very good British and European graduates in engineering and science, is a big gas turbine manufacturer located in Derby for example. Probably the government should sponsor more classic subjects though; but businesses will continue to invest in the training of people and the knowledge they think they need for the future, and that is going to be biased.
I do not agree that the sole reason for the difference in funding are the labs. Social scientists will require travel, will also use supercomputers... It is also possible to do the hardest of sciences, say in math, with a pen and paper. In general, yes, engineering physics and bio-sciences will require big hardware, but the reason for the difference in funding is elsewhere in my views.