Pedantic point: not everything is Gaussian. Income, for example, doesn't go below zero. It may not even be log-normal, but rather a power distribution.
Pedantic point aside, you bring up an important point. Can behavioral incentives be aligned to avoid the bottom side of participation?
This is less of a pedantic point than might be assumed.
The Gaussian depends upon a notion of independence and identical distribution: independence can be dispensed of when one thinks about power laws.
If you can soberly state the assumption that the effects of a comment on any other comment is completely negligible, or falls into a couple of other strange exceptions that leads to the universal phenomena that generates the Gaussian, then and only then can you state that any quantity associated with comments is Gaussian. I do not think that this can be soberly stated as such.
If you measure income as change in net worth minus expenditures, it could go below zero. That might not be an appropriate definition in all cases, but, for example if you incur debts without acquiring an offsetting benefit (maybe by being compelled to pay damages to someone for something you did that you didn't derive much or any value from, like causing an accident?), it might be reasonable to say that you had negative income. Or maybe due to capital losses -- you own something that gets stolen or destroyed or damaged, or whose market value falls a lot.
(That doesn't mean that it will follow a Gaussian distribution, just that it could be defined in a way where it makes sense that it can sometimes be negative.)
Pedantic point aside, you bring up an important point. Can behavioral incentives be aligned to avoid the bottom side of participation?