It acknowledges that achieving a certain behavior—whereby transactions once evicted cannot re-enter the mempool immediately—might not be straightforward without compromising another desirable property. This raises questions about whether the pursuit of an optimal solution is necessary or if incremental improvements should be considered sufficient.

Further, the difficulty in deciding which child transaction to evict when faced with multiple candidates is highlighted. The sender suggests that striving for only the optimal evictions might be impractical and that a more flexible approach could still yield benefits. They propose to put aside the quest for perfection in favor of addressing the concerns in simpler scenarios first.

Additionally, an alternative strategy is revisited, which involves allowing temporary breaches of cluster limits followed by a process of linearization and pruning to meet the established limits again. This alternative is critiqued for potentially being limited by the practical challenges of reducing the size of sibling eviction candidates. Despite previous dismissal, the sender argues for the importance of considering all alternatives collectively for a comprehensive understanding of potential solutions.