It surely can't be right that a teacher or care worker or research scientist is expected to pay the same graduate contribution as a top commercial lawyer or surgeon or City analyst whose graduate premium is so much bigger.
But they're already making a greater contribution - they pay a huge amount more in tax. This reflects the fact that we actually want people doing degrees in law, medicine and for a city analyst, I'd guess math (if we're talking quants). These people are not a problem, and we want more people doing those sort of degrees precisely because we actually want those sorts of skills.
So, the result of Cable's idea is that we don't deter people from doing degrees where there is little demand for the resulting education (Philosophy, Sociology). We still incur the costs of worthless degrees. On the other hand, we're going to encourage people with the sort of skills/abilities who earn lots of money and pay lots of tax to go and work elsewhere.
Here's a better idea: lower targets about higher education and drop the amount of spending to an expectation of less than 10% of people going to university. The best of the best will go to university. We just don't need 40 or 50% of people with degrees. Take computing... the likes of Google, Microsoft and Pixar need people with comp sci degrees and PhDs, but most corporate programming doesn't. A degree has only become a requirement because of the number of people with degrees.
Dropping the spending would focus universities on the courses that they really need to provide rather than the slurry degrees they do now.