I've analysed the BITRE data between August 2014 and August 2016, and in summary I think the numbers do add up for a PER-LHR flight. Note this is based purely off the BITRE numbers - it does not take into account passengers having stopovers (they are considered by BITRE to have terminated their journey in DXB) nor EK feed onto the QF DXB-LHR legs.
Firstly, 54% of QF Australia - London traffic is carried on QF 1 (the rest on QF 9). This meant on average there were 228.6 passengers per day on QF 1 travelling directly to London - note the closeness to the capacity of the 789. Meanwhile on the inbound journeys, 60% of QF LHR-Australia passengers are on QF 2 which can be partly attributed to the greater connection possibilities arising from QF 2's arrival time in Australia than QF 10. Not surprisingly, there are 147 passengers per day on QF 10 flying LHR-MEL on average. For reference, the monthly average of loads on QF flights between Australia and DXB was approx. 76%.
What's interesting to note is that collectively 60% of all passengers on QF 1 & 9 continued on to London on the same flight, the rest either terminating or connecting in DXB. On the return it was closer to a 50:50 split between those who originated in LHR and those who joined QF 2 or 10 in DXB. When considering the connection possibilities at DXB, both QF 2 and 10 can connect with most flights from EK's European destinations. QF 9 connects with EK's largest European departure bank, but QF 1 only has connections to only 5 of EK's European destinations out of 37. This suggests to me that there is may be a sizeable amount of QF 1 passengers that currently backtrack from London to their final destination.
This leads me to create an example schedule when PER-LHR comes online.
QF001 SYD1630 - 1830PER2000 - 0700+1LHR 789
QF002 LHR1200 - 1400+1PER1530+1 - 2230+1SYD 789
QF009 MEL2330 - 0630+1DXB0830+1 - 1240+1LHR 388
QF010 LHR2115 - 0815+1DXB1000+1 - 0630+2MEL 388
QF031 SYD2230 - 0600+1DXB 388
QF032 DXB1000 - 0645+1SYD 388
IMO, the combination of 1x 789 and 1x A380 better matches the LHR demand. There are also some other benefits: the proposed lunch time departure from LHR has more connection opportunities in PER than QF 10 currently does in MEL, hopefully increasing actual passenger numbers; F is still available SYD-LHR; more European connections in DXB for those travelling on QF metal from SYD; and the re-timed QF 10 offers more connections in MEL e.g. to Tasmania or regional Victoria.
Sir Tim Clark has stated that PER-LHR is the "smart thing to do" (
https://au.news.yahoo.com/thewest/wa/a/32871043/emirates-boss-backs-perth-london-non-stop/#page1). While I wouldn't read too much into that, it shows that EK is aware of this development and will probably take it into account when the JV is renegotiated. Both airlines have stated that they wish to extend the partnership beyond 2018.
In my hypothetical schedule above, you can see that the 789 would originate from SYD, which has a good likelihood of becoming a 789 base. It would not surprise me if PER was to become a scissor hub for QF metal (carbon?) flights to Europe in the long term a la SIN pre-EK days, with all other European destinations to be reached via EK from DXB. I am not going to argue whether point to point travel is economically feasible or not, but PER-LHR will certainly not be a point to point route in this scenario.