Depends on how several factors: how "bad" the review is, what other considerations exist for the device, etc. Two related examples, both reviewed and discussed fairly recently on ASR: the NAD C 658 streaming/DAC/preamp and the BlueSound Node (same parent company, both use the same BluOS software).
After the NAD review, we did not see multiple posts telling about people selling their units. Why? Because even though the measured performance was mediocre relative to the best available, for most music in most systems for most people, any differences were likely not audible. On the other hand, the 658 has a huge array of useful features and mature control software. Had the performance been abysmal, the story might be different, but sometime "good enough" is indeed good enough.
In the case of the Node, it appears that most of the people who commented were using it as a streamer and bypassing the DAC, i.e.,using an external DAC. Thus, the measured performance that included the DAC was irrelevant; Amir's measurement when the internal DAC was bypassed were fine, i.e., limited mainly by the stream source. Again, the features offered outweighed the less-than-state of the art measurements. And even those using the internal DAC again thought that the performance was "good enough".
BTW, saying "good enough" sounds like faint praise, but why worry about issues that are, in most cases, inaudible? That would not be very scientific!
