Good advice, and it would be my action as well when dealing with this rather poor opamp layout works by Yamaha.
There is, however, a mountain of nuance on this topic.
Some default reading to do:
http://www.analog.com/media/en/training-seminars/tutorials/MT-101.pdf
and for high-speed opamps (rather beyond analogue audio applications, but informative nevertheless)
http://www.ti.com/lit/an/sloa069/sloa069.pdf
In addition a couple of text books on that matter, leads me to rather use NP0/C0G MLCC capacitors than "film caps" for reasons of significant higher ESL for film caps.
Secondary, such decoupling is bound to very short distance from the opamp pin to a "very low inductive ground plane" to keep inductive elements to a bare minium (or else resonance and noise are rather generated)
And that might be the second issue for CX1000: the + and - rails are readily under the opamp of course.... but where is the ground plane (0V)?
When long leads are required to connect the decouple caps from power rails to ground, then there is a risk of adding additional induction to the solution.
So yes, the Yamaha PCB layout is piss poor regarding opamp decoupling, but it might be a tricky treat to improve it afterwards.
A lot of care and knowledge required to do something that makes sense.
EDIT: the power supply decouple caps should be between rails and ground; NOT between +/- rails !!