First of all, I understand that flag weight is not like reputation, and in the grand scheme of things, you can participate perfectly well in the community regardless of your flag weight, but theres something that's been itching me for a while now with the way flags are sometimes handled.
See How to know if an app runs for the first time on device? on Stack Overflow. It is clearly a duplicate of Android Apllication first run. It is so clearly a duplicate that the accepted answers are exactly the same, albeit with vastly differing amounts of detail. I flagged the latest one as a duplicate but the flag was declined and the question is still open.
Now, see How does Stack Overflow create its question URLs? I flagged as off-topic to be migrated to meta (admittedly, maybe a little premature) and the flag was declined, but the question was later migrated to meta anyway.
I flagged an answer for Android Facebook API errors with SSO when Facebook App installed as not an answer, which was declined and then the answer was deleted.
I think we need to consider a better way of handling flagged questions or answers in situations like this.
I'm usually an enthusiastic flagger, and if I make a bad call (like the flag to migrate I mentioned above), then I'm happy to take a hit to my flag weight but when my flags are declined, and then later the post is flagged by other users and acted on by mods for the exact same reason, it's a bit of a kick in the teeth when you are actively trying to improve the site and are losing flag weight when you are subsequently proven correct.
I know the arguments for leaving the system as it is, but the fact is that flag weight is used to determine how prominent flags are in the moderator's queue and therefore have a direct, tangible impact on the quality of the site as a whole.
I'm not asking for flag weight to be awarded for flags that are subsequently approved (though that would be nice, especially when identical flags from other users are deemed valid/helpful), but can we at least treat them as disputed and remove the flag weight penalty when the same flag by different users is deemed helpful?
I also understand that this is a relative edge-case, but considering the potential good you can do with flagging it seems to me that these edge cases cost the community enough in terms of quality of questions and answers that a relatively narrowly defined "flag recalc" to remove a "bad" flag penalty would help improve flagging (and therefore directly improve the overall quality of the site) overall.