So say they throw out 5% of the results or give a +/- rate you can measure to the final stats and use that to determine whether or not it was significant enough to warrant a recount or a new measurement. If not then it's probably not worth the time as you can assume the majority party will be satisfied with winning and the losing party would not benefit from a recount.
Additionally, the bar would have to be extremely high that the likelyhood it coulld affect the outcome of the results given the extremely high cost involved in doing the data collection or analysis over again.
So, basically, the OP made a fair point.
Whether or not society should be okay with the level of irregularity and fraud is another question. Even if it didn't ultimately affect the results, that doesn't change the morality or questions of legal responsibility of those who manipulated the results.
• There were emails in the DNC leaks that hinted at a suspiciously cozy relationship between the party and media organizations.
• During the DNC primary process, media reports often included superdelegates in Hillary's delegate count, making Bernie's case seem impossible. Media often failed to even mention the existence of Bernie Sanders in many cases, or calling him "unelectable" even while he was polling ahead of Hillary or Trump.
• In prior elections, some news channels may have gone as far as not even listing a second-place candidate they don't like in poll numbers.
• Voters can be influenced by a belief that their vote will be wasted by voting for someone who is not a "sure thing".
• Therefore, (assuming) the media manipulation or incompetence, DNC bias in favor of Hillary, and vote manipulation to give the impression of momentum to Hillary, even an insufficient number of delegates won by "cheating" could have swayed the election by second-order effects.