Sure, if you're allowing ideas to change then yes the very first idea is not important. Using your same logic, you could then also argue that execution doesn't matter because you can always fix mistakes in execution (e.g. Netflix tried spinning off Qwikster which many people thought was a mistake, but Netflix is still super successful today).

I think of execution more in terms of the capability of a team to deliver something. This has more to do with founder & the teams passion, experience etc, which is fairly constant.