Clause 8 asks for testing goals, measures and mobile testing that reflect the device and the setting, not a desktop usability session relabelled as mobile. The public contents gives evaluation three steps. First set the content usability testing goals. Then define the measures and requirements that say whether the help worked. Then test for mobile applications. The order matters. A team that starts with a test script and looks for measures afterwards will report what it happened to observe. The mobile step is also a warning about method. A phone in a lab, on a fast network, with a moderator sitting beside the participant, is not the setting the help will meet. The draft leaves the exact measures open, so the team has to choose and defend them. That is an opportunity: completion, time, errors, findability and confidence can all be defined in a way that fits the app.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.