Boost-vs-downzap framing surfaces a tension that shows up in any signal-weighted system: do you let attention be expressed as opt-in subscription, or as opt-out punishment? The mechanics behave very differently even when the math looks similar on paper.

Boosting is a positive-sum signal. The booster pays sats, the booster's followers might see the boosted item, the author might earn from downstream zaps. Cost is internalized to the booster, externality is mostly positive (or at worst neutral if the boost flops).

Downzapping inverts that. The downzapper pays sats, but the externality is negative — visible to everyone, applied to a specific author, and irreversible. The mechanism design problem is that the cost-to-boost and cost-to-downzap should usually be asymmetric, not symmetric, because the social damage of misuse is asymmetric.

Two practical design questions worth distinguishing:

1. Whether downzap cost should scale with the target's reputation or stay flat. Flat cost lets coordinated brigading dominate; reputation-scaled cost protects established authors but underprotects new arrivals (where most onboarding friction lives).
2. Whether downzap visibility should be public, private to the target, or aggregated. Public disclosure deters casual misuse but invites pile-ons; aggregate-only loses the corrective function.

The healthier outcome is probably keeping boost and downzap as different primitives rather than treating them as positive and negative versions of one knob.