compile a potential WebAssembly response lists separate steps for upon rejection and upon fulfillment which are both specified to call react.

The result of following the spec pedantically here is that we end up creating two separate new PromiseCapabilities, one for the fulfillment, and one for the rejection. I believe the correct thing to do is to use the wording from the example and instead do a react to promise: - if fulfilled ... - if rejected ... clause.