This isn't always possible — for small apps, sure. But for apps with millions of users (many of which have tiny / no support team), the nth user report isn't helpful and ends up taking time away from fixing the issue.
If it's a Microsoft app, I will rate it based on its performance at the time of the latest release. They should know better, and have more resources to catch these things before release. In the example I cited, I was willing to bet the author didn't have a WP8 device to test on, and I was using the app problem free on WP7. Microsoft or other giant publishers should have these resources, you are correct.
