Furthermore, two-party Payjoin faces a significant privacy limitation: because only two peers interact, they still each know each other's inputs and outputs. This means that while Payjoin provides privacy benefits against third-party observers, the participants themselves have complete visibility into each other's transaction details.

Multiparty Payjoin solves this "second-party privacy" issue inherent in the two-party model. By involving multiple participants, they might build transactions where no single party has complete knowledge of all inputs and outputs, thereby preserving privacy not just against blockchain observers but amongst the transaction participants themselves.

Payjoin V3 will solve these limitations by introducing a collaborative model where multiple parties can contribute to a single transaction. This creates a flexible network of participants all working together to build more efficient transactions.

Unlike previous versions where transactions followed a rigid request-response pattern, V3 will allow transactions to be built collaboratively with multiple participants adding their inputs and outputs to create truly optimized Bitcoin transactions. This approach effectively enables Greg Maxwell's transaction cut-through concept originally proposed in 2013, which suggested that Bitcoin transactions could be combined to improve privacy and efficiency on the network.