Checking the requests from the webserver proved problematic, as often there's
some kind of caching going on. Instead, we only check the log, as this is used
for things like :navigate anyways, so if the log says the page got loaded, we
can trust it.
There's still "... should be requested" to check the actual requests.