In their comparison, Dan and Justin used the 30 day t-bills for the cash component which is as literal as you can get with cash.
Personally I'd hold something like XSB as a short term bond fund (Americans use SHY which is similar to XSB). This is a common adjustment to the permanent portfolio, and you'd get some increase in performance. Quite a bit actually. XSB since inception returned 4.5%
Assuming that this XSB performance since inception is indicative, you gain another 1.1% performance in the permanent portfolio. Which, surprise surprise, now makes the Permanent Portfolio have equal performance to the Global Couch Potato portfolio. (let's call that the optimistic result)
I really don't think the PP performance is bad at all, and the XSB substitution is a pretty reasonable implementation of it. I'm not convinced that you sacrifice much performance with PP.
Even if I'm wrong on the XSB performance projected back to 1979, let's say that 1-3 year bonds definitely yield around 100 basis points more than 30 day bills (pessimistically) which closes the gap in the comparison by 0.25%. In any case, the message is that the performance gap is narrower than the 1.32% figure in their analysis and the long term annual PP performance is probably within 1% of the 60/40 couch potato.