A growing body of literature is aimed at designing private mempools in blockchains. The ultimate goal of this research is addressing several phenomena broadly classed under MEV with sandwich attacks as the canonical example. The literature has primarily viewed MEV as a problem arising from oversights in distributed systems and cryptographic protocol design and has attempted to address it with the standard tool sets from those disciplines. This paper argues that the impact of private mempools on markets and agent incentives renders analyses that do not consider the economic lens incomplete. The paper presents several observations across blockchains and traditional finance to justify this argument and highlight specific dynamics for future study.