I consider your analogy inaccurate and offensive.
I will discuss the medical company part first.
I think that you may have a misunderstanding about how drugs (and implanted medical devices) are approved for widespread clinical use. What I say next is a gross over-simplification with some attempt to make it a layman explanation. In particular for investigational new drugs in the United States, this is a FDA-regulated multistep process that considers relative efficacy and risks associated with the investigational medicine versus typically the current standard of care or in combination with current treatment regimens. New drugs are not approved because they work on an outlier individual. Certainly, a drug does not need to work on all patients but can be a subpopulation (e.g Herceptin in HER2-positive breast cancer), based on personalized medicine and molecular characteristics.
If a drug only shows evidence of efficacy in an outlier, that drug would never be FDA-approved. A important concern that comes up sometimes in this process is that during the clinical trial, a pharmaceutical company may change the clinical outcome measure to demonstrate efficacy after the fact or do a post hoc test to define the population where the drug may have efficacy but have neglected Bonferroni considerations.
In general, lack of efficacy is not what gets pharmaceutical or medical device companies sued. The primary reason relates to adverse events. All medications are associated with risk including simple over the counter medications such as aspirin. Implantable medical devices can fail and drugs can have severe side effects, causing harm and/or death. During clinical trials, there are mechanisms in place to report adverse events and the clinical trial lead is required to report all adverse events. Moreover, pharmaceutical companies are supposed to follow their drugs after FDA-approval to see if in a larger population adverse events are occurring. As you can imagine, this does create a potential conflict-of-interest as the adverse event reporters may have a financial stake in the success of the trial. Safeguards are in place to prevent this misconduct, but no system is perfect. For example, all doctors can report adverse events that they think are caused by a particular medicine. So why are pharmaceutical companies being sued? Typically, it relates to underreporting adverse events or when adverse events are reported that they do not act quickly enough to determine if their drug is causative. As an epidemiologist will tell you, it sometimes can be rather difficult to show causality in human population given out genetic and environmental heterogeneity.
You quoted point #1. You did not quote point #2 in that post (#91) in your linked thread. I explained in more detail later in that thread about interpretations. My analysis does not rely on an outlier for its conclusions. In the PTR, there is 3X the number of players who cleared a GR126 with barbs than the numbers of players who did so in the current non-season era using demon hunters. The top DH, monk, and crusader, in non-season in the America/EU/Asia regions have not cleared a GR140, but a barb within 24 hours of the PTR going live did.
If you read this thread thoroughly, you will realize that I strongly disagree with the opinions of several poster. However, I think it is fair to say that there is an emerging majority who feel for comparing relative power, it is best to compare classes using non-season/era numbers. Comparing a seasonal clear to non-season has a confounder of the seasonal buff. Also, we agree that you can not compare clears using only PTR numbers, due to bias in the number of players testing each class.
er, just because Barbs could suddenly be 2 GR or max 4GRs ahead from other classes…
X 100 That will have to do. 
