I'm not a software guy but there is a lesson in technical program management playing out IMHO.
There is a discussion over what kind of replay protection to employ:
https://github.com/BTCGPU/BTCGPU/pull/62The meta point is about how much extra technical risk do you accept at the end of a tightly scheduled effort?
If other engineers are telling you that what your doing is going to balloon the effort's schedule and interfere with adoption, it's worth seriously listening to.
If you are cheaping out and staffing lean, then the management consequences are that you have to be more serious about outside input.
This is for right now a BTCGPU implementation speedbump.
