Cafe City
Glad you like the SIKOU reversal, only took 7 years from planning to implementation. Can't really comment on the lack of talk on the first 2 freqs. inbound but sounds like things were going smoothly.
The problem with sequencing into HK is that there are basically 7 streams of inbound traffic being integrated into 1 via 3 intermediate stages - ABBEY, BAKER and MURRY. The first 2 are relatively easy to set-up with an in-trail spacing (10nm for a single feed, 20nm for a dual feed). The third MURRY has its own problem in that traffic via SIERA can't be integrated into this neat in-trail scenario until it hits SIERA. That's where the trouble is, in that with 3 gate feeds the spacing needs to be 30nm in-trail through each one. Ideally, if HK had control of the SIERA traffic from top of descent then it could be integrated into the other 2 streams more smoothly.
Why 10nm/20nm/30nm? Well, the average speed through ABBEY/BAKER/SIERA should be reducing to about 280kts which then compresses down to about half that speed on short final as does the spacing. The aim being to have about 5nm spacing on final. HK has the overall picture but it doesn't have the overall control of all the inbound traffic until the SIERA arrivals join the fray. This is where rapid integration has to take place and will invariably lead to what may seem like a chaotic string of instructions. Unfortunately, things are unlikely to change until the whole Pearl River Delta traffic situation is rationalised and HK gets to control SIERA traffic a lot earlier.
If you can, I would urge you to pay a visit to the ATC centre and see the whole picture for yourself. Then, I think you will be able to see exactly what I'm on about and the guys there will certainly be able to explain it better than my attempts here.
On the beach
ps. I like the moniker