Strictly from a dollar-and-cents point of view, curation will be more profitable if and only if the content that you curate gets a lot of other votes after yours.

One issue here is the approximately 80% cut (50% nominal cut, plus a larger cut due to the 'reverse auction') that was made to curation rewards earlier on, a cut that I opposed. Instead of curation and content being equal in value (a balance that would make earning the same or greater amount via curation much easier, though still not guaranteed), content is _heavily_ favored. 

In slashing curation to almost nothing, it made other 'alternative uses' for votes more attractive. Previously those alternative uses were things like whales hiring or partnering with writers, voting for them, and getting a lot of the rewards back under the table. Now that non-whale votes have real value, it includes (non-whale) self-voting as well.

The underlying problem is that real curation has insufficient incentives (in large part due to the 80% cut).

Still, in cases where you can recognize good content early, you can probably make more with curation, but the way the reward share has been set up, the system is unfortunately working against this sort of desired behavior.