Thank you for your insightful feedback, Craig. I wholeheartedly appreciate the points you’ve raised, particularly concerning the challenges faced by MPTCP in environments with fluctuating WAN link conditions. Your observations about the necessity for closely matched performance characteristics in bonded lines are indeed crucial for mission-critical systems.

While I recognize the limitations of MPTCP when deployed in isolation, I’d like to highlight the efforts by OpenMPTCProuter in addressing some of these challenges through configurable routing algorithms. Their “Redundant TCP Scheduler,” for instance, offers a noteworthy compromise between stability and speed. It’s tailored to improve reliability, especially when augmented with a v2ray proxy and SQM adjustments capped at 300ms. These measures have shown to maintain acceptable performance levels for applications sensitive to latency, such as video calls. There’s a detailed case study shared by an OpenMPTCProuter user that delves into these adjustments further, which can be found here: https://github.com/Ysurac/openmptcprouter/discussions/2594.

Indeed, relying solely on MPTCP might not meet the rigorous demands of certain applications. However, the suite of solutions provided by OpenMPTCProuter presents a viable, mostly-free, self-hosted DIY alternative for those looking to mitigate some of the inherent challenges.

Nevertheless, for professional and mission-critical environments where reliability cannot be compromised, the specialized solutions offered by your company, such as Broadband Bonding Network Appliances (BBNA), represent a superior choice. Your products’ ability to deliver robust performance under varying network conditions is commendable, and I frequently recommend Mushroom Networks in such contexts.

Thank you once again for engaging in this discussion and for providing your valuable perspective.