How should Audio occlusion be applied?

I am developing a third-person multiplayer game using Unreal 4.72.2 / FMOD Studio and Resonance Audio.

I am experiencing issues with applying audio occlusion.

Current method: It determines only the presence of obstacles between the sound and the listener. If there is an obstacle, occlusion is applied; otherwise, it is not applied (no variation based on thickness or material).

Improved method: When an obstacle is detected between the sound and the listener, measure the thickness of the obstacle and apply occlusion variably based on that thickness (apply based only on thickness, excluding material).
For thin obstacles (e.g., wooden doors), if there are two, determine the distance between obstacles as the thickness and apply occlusion.

While testing in levels, there are occasional instances where excessive occlusion values are applied even when the thickness is not substantial. (This phenomenon mainly occurs with pillars, large rocks, or when there are other levels involved.)

Currently using the Visibility trace channel, I am considering creating another channel.

I am debating whether to continue using the current method or adopt the improved method with a new trace channel. Alternatively, if you are aware of another method, could you please share it?

Thank you in advance.