Don't know about 20+ frames for just 6 routes. Qantas keeps 2.5 planes for each of its circular long-haul routes and 2.2 for co-located routes, such as LAX to BNE/SYD/MEL, where you only need 1 spare plane at LAX in case things get sour with maintenance (not to mention the one docked at JFK in absolute emergencies), and one spare at Sydney, which is how they currently run them. That's why my bet for Qantas was 10-15 frames, depending on whether or not BNE/MEL-JFK on their own are actually viable (and maybe BNE feeds MEL or vice-versa to make 2 flights viable). Can't much complain about 1 stopover in Sydney/Melbourne when it comes down to economics, and the layovers would be very short.
Like I say, we have been over this before. If you think we are only talking about six routes you haven't been following what Joyce has been saying. Yes, there will be a handful of ULR sectors (LHR, NYC, ORD, possibly CDG, FRA, and eventually YYZ, MIA). And there will be some very long sectors that would be better served by a high-weight 35K than an A380 (DFW, GRU/GIG). Plus a bunch of Asian destinations – Joyce hasn't said which ones but he has
said that they will use the Sunrise fleet for those routes as well as for long haul. You can make your own guess about which of these will be served from SYD, MEL, BNE and PER. And yes, some might continue to be served by the 789.
But you need to understand that Joyce is looking at a big fleet of a single type, presumably with a sub-fleet (at least seven) of the ULR variant. Single pilot pool, common maintenance and training, interchangeability across the fleet for most purposes. Run the numbers. You can easily get to 20+ with plausible assumptions about which destinations will be served from where.
And my wonder in all of this is why Boeing and GE haven't considered pushing an NG model of the 787 to meet the job. The GE9X has been providing the GEnx lots of PIP items. I wonder if they can scale the 9X down somewhat and keep the efficiency. That and a couple tweaks to the 789's frame and tanks (maybe folding wingtips on a longer wing?) would probably be an excellent option, as it means fewer types in Qantas' fleet, good for costs all around.
And my wonder in all of this is why so many a.netters are stuck in fantasy mode on anything 787-related. Yes, we will see a neo version of the 787 at some point. But right now there is no new engine and Boeing have their hands full with the MAX, 77X and NMA. We don't even have confirmation there are engineering mods on the 78Js that NZ ordered. The fact that the NZ CEO said they will operate mainly on Asian routes suggests that any near-term improvements on the 787 are going to be minor.