I have seen many simple questions (and even answers) with a very large amount of up-votes. Is it a good idea to give so much reputation just for a very simple question or answer which is just because it is a common request of people who search the web for it (and the question/answer author is one who just posted it sooner than other ones)?
As an example, consider this simple (one line) question on Stack Overflow:
Q: How do I redirect to another page? link => 6041 up-votes!
A: window.location.replace(...), with a little more explanation link => 10008 up-votes!
Also consider a 10 points gain for each up-vote! it could theoretically give authors 30k and 100k reputation! (If distributed per different days in 2 years, but, anyway they will produce a very large rep despite the per day limit.)
What is wrong with this? I agree that the up-votes should be saved in the system and also shown for each post (to show how popular the question/answer is), but the reputation assigned to the user can be controlled such that they don't get unnecessary points which other users should try hard to achieve.
So, to improve the reputation system, I think there must be some limits for the total reputation one user can get from each question (or answer), or some weight factor for votes higher than some limit.
Consider this with the fact that many good technical questions and answers provide a very low reputation gain (e.g., +15 for accepted answer or if you have chance for 10 upvotes: 10*10 = 100), or consider many hard questions suggest a bounty worth about +50 or +200 rep. How could we compare them?!
As I already mentioned, I know there is a limit of 200 rep for each user per day. I think its not enough to control the issue as I've explained above.
I don't say we should distinguish simple and non-simple q/a. I think that a limitation or weight on reputation per post will improve the meaning of the reputation system, better than what is now in place.