Hot switching kills RF switches slowly, and datasheet lifetime numbers assume ideal conditions. Has anyone measured hot-switching lifetime under realistic VSWR and power?