The key point which you flew by is that a linear split is not an effective method of gaming the reward pool, as long as it allows free entry, and continues to have support from larger stakeholders. Any small stakeholder can enter as free rider which destroys the economic value of the scheme (if it were a scheme in the first place) for anything other than redistribution.

When it comes down to it though, "sybil attacks" (arguable whether to even call it that) by (smart) whales probably won't be recognizable or likely detectable at all. The fact that large stakeholders can direct the reward pool to (direct and indirect) benefit is an almost inevitable consequence of the entire design of the platform (i.e. it is by design). Only if the larger share of the voting base agrees that rewarding blog posts is a good use of dilution (i.e. increases the value of the post-dilution stake) will it actually happen. There is no way to force the voters' hands.