When writing a proposal, the most egregious mistake you can make is asking for something that's already been proposed. Not only are you contributing to the noise, but you may very well be splitting the vote for your request, making it less likely to be accepted.
✓ Find existing proposals for your problem.
In a perfect world Linden Lab would notice and combine all votes for a particular feature. With so many proposals, Linden Lab will surely look first at the proposals with the most votes. If the total votes to solve a particular problem are split among too many proposals, the community's actual desire for a solution is not readily apparent. Your vote could be overlooked.
As an experiment, I made a list of the top 100 proposals by vote count, and tried to collapse multiple proposals for the same feature by hand. Here are the top 15 desired features by vote. Note that while there are five collapsed features listed here, only two of those features were represented in the top 15 before I collapsed them.
|2||Try Before Buy||1590|
|3||Allow multi-select in inventory||1405|
|4||Point-to-point teleporting (32, 89)||1344|
|5||Bill of Rights||1281|
|6||Hide from friends (52, 474, 479)||1140|
|7||Generalized Texture Layers||1022|
|8||Better/custom default animations (67, 25)||1011|
|9||Bring back event support for events other then education||993|
|10||Native Linux Client||922|
|11||Object to Object communication||790|
|12||Non-primitive building primitives (49, 318, 138)||781|
|13||Real Time Weather||714|
|14||Mature-only SL (279, 459)||624|
|15||Register Mafias-No Report Rule||556|
If the creators of duplicate proposals had looked for existing proposals that matched their desires, so many issues would not be so ill represented.
While looking for proposals that address your problem, you may find some that come close but don't solve quite the same problem, or don't address an aspect of the problem that's important to you. You may notice that I combined proposals #67 User-Defineable, Replaceable Default Animations and #25 Improve Avatar default animations above. These proposals both deal with the stock avatar animations, but suggest very different solutions: #25 is simply that the stock animations should be replaced with new and better animations, whereas #67 is that residents should be able to select what animations play for the stock animations' events (that is, build animation overrides into SL). While these proposals differ, they both address one problem, namely deficient stock animations.
My advice for residents in this situation is, if at all possible, get in touch with the author of the other proposal and craft a new, combined proposal. The new proposal should address the problem, and suggesting either or both of your solutions as potential remedies. For example:
#600 Fix default animations
The current default animations are not very nice. They could be improved in two complementary ways:
1. Replace the current stock animations with new ones. This would be an opportunity for an animator contest, in the vein of the game developer contest and avatar expo.
2. Enable residents to select animations in their inventories to replace default animations. While this would kill the market for animation override devices, it would encourage new innovation in animations. Default animation selection could be an extension of the existing Gesture system if need be.
So what can be done now, with so many duplicate proposals already filed? In addition to direct action on the part of proposal authors as above, Linden Lab might consider adding a notification system to voting. The system would allow the resident who opened a proposal to notify all who voted for it that there's a better game in town. For example, if someone added a new proposal called "Free pony rides," I might be able to select an option on the page for One (1) pony to send a note to all voters suggesting they confer their votes to "Free pony rides" instead.
This vaguely resembles logrolling, the age old political practice of you-scratch-my-back. Logrolling differs in an important way though: where it's a two-way trade of votes for two issues, a vote consolidation system transfers votes for one single issue to someone else's similar proposal. It would be interesting to consider if such a system could enable actual influence peddling, though.
Until Linden Lab adds such a feature to the voting system (and even then), you should always check for existing proposals that already ask for the thing you want, and simply vote for that proposal instead of starting your own.