These observations should be taken with a grain of salt. They are the views of one lawyer/mediator and thus are subjective and necessarily based on a limited sample size. They may simply reflect the views of an older person. To quote Bob Dylan: "I was so much older then/I'm younger than that now."
With that said, it seems to me that there is more uncertainty regarding the outcome of commercial litigation now than there was 20 years ago. The change is subtle and varies among jurisdictions.
In the federal court system, judges seem much more willing to decide cases on summary judgment than they were even ten years ago. I have read more and more decisions where federal judges essentially just decide the entire case (including damages) on summary judgment even though the case seems to have rather obvious issues of fact that should be tried.
On the one hand, some would argue that this adds to certainty, as it means the case is not put in the hands of a six person jury. On the other, it means that one person is going to decide the entire case.
Federal judges tend to be well-qualified and experienced, and they try to get it right. However, all judges are human and do not always decide cases correctly. In the insurance area (one of my primary practice areas), there are countless examples of federal judges making incorrect guesses regarding state insurance law. (For the non-lawyer readers, federal courts deciding state law questions are required to apply the law of the state in which they sit, and may have to guess how the state courts would decide if there is no clear state court precedent).
I am currently writing a law review article on this subject, but my article on insurance coverage for construction defects covers one area where the federal courts repeatedly incorrectly applied Georgia law until corrected by the Supreme Court of Georgia. There are many other examples.
If you think you are protected by an appeal, think again. Reversals in the federal system are becoming very rare. Although federal appellate courts are supposed to review cases granting summary judgment de novo (meaning to take a fresh look at the issue), this now seems much more theory than practice.
In the state court system, results are also uncertain. Most state court trial judges carry enormous case loads, and have had to deal with staff cuts. Cases involving hard legal issues merit careful thought and study, but the practical realities can intrude.
In Georgia where I practice, appellate review in the state court system is more likely to result in a reversal than in the federal system. However, the results are still not predictable.
The legal profession must share some of the blame. Throughout most of my career, it was standard practice for lawyers to conduct a thorough legal analysis of all but the most routine issues before filing a case or an answer. Today, clients rarely want to pay for this even though, in reality, a thorough legal analysis almost always results in better strategic decisions and better results.
Many lawyers thus file briefs that do not adequately address the issues, and overworked judges with few resources decide cases without the benefit of being provided sound legal analysis and sometimes even citations to controlling precedent. Westlaw and Lexis, the two large providers of legal databases, now make access to the parties' briefs in cases accessible. Thus, when one runs across a questionable court decision, it is often possible to review the briefs filed by the parties in advance of that decision. There are instances where questionable decisions seem to have resulted from the lawyers simply not making the correct arguments. Unfortunately, the resulting decisions become precedent, which only adds to the uncertainty.
What does all this mean for mediation? If a lawyer or a party is absolutely certain about its legal position, it should be very careful. There are certain cases where there is a controlling precedent directly on point, but they are relatively rare. Other than such cases, the chances of a court getting it "right" (from your point of view) are probably less certain than in the past.
Thus, when a mediator suggests that the result may be less certain than you may think, pay attention. The "twenty percent rule" may be creeping closer to the "twenty-five percent rule."