Starting the 1980s, extremely powerful computational fluid dynamics software has been used to design wing sections and whole wings. I am going to focus on the Boeing 777 as an example of CFD applied to commercial airliner design since this is the aircraft that Boeing itself claims had an enormous use of CFD in design of the wing and wing-airframe integration (source). To quote, "The 777, being a new design, allowed designers substantial freedom to exploit the advances in CFD and aerodynamics. High-speed cruise wing design and propulsion/airframe integration consumed the bulk of the CFD applications". In particular, "inverse design" was first used (at Boeing at least) for the design of the 777.
Conceptually, CFD is a tool in the process of wing design, a tool that must improve attributes of the wing that ultimately then result in improvements in parameters which customers (airliners) care about. For example, one imagine the following causal chain: better CFD predictions of wing pressures --> some changed wing characteristics --> lower drag for a given lift --> lower fuel burn (which airlines care about).
My question centers around the "changed wing characteristics" part of the casual chain above. That is, I'm unsure as to what specific wing characteristics changed as a result of the application of CFD in the process of wing design. This could be changes in wing planform, thickness, t/c tapering, twist, or something else entirely. Basically, there has to be some output of the application of CFD that shows up in the actual wing itself to make the wing more efficient, and I'm curious what that output was.
Note: edited for clarity.