Ah that brings back memories. RTR Phillips and Malcolm Bird. That article dates from early 97 at a guess, when DNA&FP thought the 40 aircraft ship was a 40000 tonner powered by WR21 (lucky escape!). At that time, people still thought a STOVL ship would be significantly smaller than a CTOL ship -one of the main reasons for the assumed "preference" for STOVL.
It predates the phase in 99-2000 where the deck management aspects were first properly considered, which identified that as you ran more sorties with more realistic packages and flypros, the size driver ceased to be aircraft launch and recovery mode and tended towards deck park size. An effect compounded when you looked at minimising deck moves between launch and recovery events, which means you don't need as many chockheads. It included quite a bit of work/cross-pol with the NavAir chaps at Lakehurst.