The sequence of event handlers' execution shouldn't be a concern. It seems that the standard doesn't specify anything on this matter. Each software implementation has the liberty to choose the algorithm that suits its workflow best.
Event handlers should operate independently. One handler must not rely on another or be affected by another's output. Ideally, each handler shouldn't even be aware of the existence of other handlers for the same event.
If the order does become important for you, then there is coupling between the handlers. This could mean that either the linkage is forced (e.g., sharing unnecessary data structures simply due to past coding decisions) and can be removed, or it indicates that the processing which should ideally be contained in one handler has been split unnaturally into multiple handlers.
In essence, your query points towards an architectural flaw in the application. By resolving this issue (ensuring independence among event handlers), the order in which they're executed will cease to be a concern. Additionally, your application will benefit from improved design, making it easier to modify and comprehend its behavior.