On Lamarr we kept the same ET specs and had a gain of 8-9%. I think that if we modified them on Crawford, we would face a lot of questions whether the yield gain is from the design or from the spec modifications. It would take a lot of engineering resources to answer the questions, but then there would still be "lingering doubts" if the yield gain was from TSR changes or from the design.
For this reason's let's simply clone the existing ET TSR for the FRL2.2B build. We can check the pareto later. I expect some specs to have close to zero loss. We can negotiate those specs with the design center.