There's a worrying lack of analysis of the actual impacts of the proposals, and how they may change behaviour.
For example, the fairly obvious fact that charging up to 35% income tax (25% flat tax relief, in the £100-125k bracket) on money you won't actually get until 20 years later is a pretty big disincentive to pension saving. People might well not have that money available! (Since they didn't get the income it's due on). This was exactly why the doctors pension fiasco was so bad.
And whilst he is very adamant that Employers pension contributions should be included in the change to add income tax - in order to avoid distorting behaviour - he doesn't mention that this effectively lowers all the income tax thresholds (which have already been frozen for so long) by a minimum of 3%, and more likely around 8%. 
He's basically a bean-counter, who only considers how the numbers affects his own bottom line, without trying to understand what the numbers actually mean.
My favourite was his outrage that the cost of the pension tax relief is increasing faster than the pension-age tax. No shit, Sherlock! It's a natural numeric consequence of increasing wealth redistribution (due to frozen tax thresholds)! And that's even without behaviour change, which happens as the frozen thresholds mean ever more people are pushed into punitive tax brackets which they mitigate through pension payments (particularly £100-125k).
And no, the best way to respond to that probably isn't (as he suggests) to increase redistribution even more to make up for the behaviour changes you've already caused by increasing redistribution...
Partway through I started thinking 'if this is the quality of thinking Labour has access to, no wonder they introduce such stupid policies'
By the end, I was despairing.
I very much hope that Labour have access to some better analysis before they make hugely significant, sweeping Pension changes.