There are many attempts afoot to establish a system of reputation to help strangers trust each other on the web, and of course which can't be gamed! Problems with conventional approaches
It is a very tricky question though because reputation is often specific to different aspects of a person's life, and specific to the context in which it arose. So the approach that tries to aggregate existing metrics gives us a lot of information but we might not know what it means. E.g.
Just heard Matthew McStravick today talking about Economy of Hours, his widening of timebanking to include local businesses.
Running the Timber Wharf Timebank in Hackney, London, he was frustrated by the limited scope of the project which, following donor priorities, had to focus a single social problem.