There is no question in my mind that an experiment must be reproducable. This does not imply that we have to reproduce it, simply that any person, any where, under simmilar conditions should get the same result. If not, either the experiment is flawed or the person checking it. If you or anyone else can't repeat what you did, what you have done means nothing. That is the whole point of science, the ablity to take something that someone else has done and aply it to what you are doing (reproducability).