The description of how user agent must handle the menu element and its contents might be overly prescriptive. For example, the current HTML5 draft specification clearly spells out when user agents should render groups of menu items inline vs. as a submenu, seemingly giving the user agent no leeway to adjust for user preference. Instead, it should acknowledge that user agents may override the default presentation described in order to comply with user preferences.

Use case: Nadia is blind and using a web browser with a screen reader. The document contains a menu structure created with the HTML5 menu element, and it includes some very long menus with many groups of menu items separated by horizontal rules into various groups or sections. As Nadia uses the down arrow key to navigate through the menu items, she has to pause for each one to be read to her, so traversing a long menu takes a long time and a lot of effort. She would prefer to have the menu presented to her in hierarchical fashion that uses progressive disclosure, so she could navigate through the short list of sections, and then through the short list of commands in the desired section, rather than through one long list of items.

Use case: Aidan is the opposite of Nadia. He uses an alternative input system and input is difficult for him, so he wants to reduce the number of actions he has to take. Therefore he prefers to see all the options visible at once so that he can choose one directly, rather than having to use mechanisms involving progressive disclosure. (He has even invested in a large, high-resolution monitor to support this work style.) Rather than choosing a sub-menu and then items from them, he'd rather have all the sub-menus and their items displayed together. Unfortunately, the HTML5 specification explicitly states that the menu element with a label must be presented as a sub-menu rather than displayed inline.