Most importantly, the criticism of a _design_ is independent of the results of the experiment.

Secondly, he may have wanted a null result.

> Most importantly, the criticism of a _design_ is independent of the results of the experiment.

Flaws can go in any direction, and make a design conservative or optimistic. In this case, the identified flaw does not undermine the actual result of the experiment - though it would have undermined the opposite (hypothetical) result of positive effects.

If you design an experiment in such a way that outcome A can be explained away by "yes, but the design is flawed", while outcome B cannot, I don't think it is fair to do the experiment.

Experiments don't have to be 'fair'; your 'fairness' is not required for tests to be informative. Imagine an IQ test which spits out a binary yes/no result based on whether you're in the top 10% or not. This is a useful, meaningful test, which is 'unfair' in your naive sense. Issues of information content, discrimination, false positives and negatives - it's more complex than saying 'it is fair to do the experiment'. In this case, I got a reliable negative, rather than a less reliable positive.