Well spotted, Warren.
This could be an oversight, but if so it is an old one as the 2002 TSRGD had the same distinction (or the 2011 No 2 amendment in the case of diag. 960.2). The 2016 TSRGD has simply replicated the previous requirement.
I’m guessing it was deliberate, as the signs indicating contra-flow cycling are larger, with greater ‘target value’ and therefore less likely to be missed.
Any other views or thoughts on this?