I see a lot of people talk about being comfortable with lower success rate percentages (sub 85%) and I don't know if I'd ever feel comfortable with levels that low.
I'll use the analogy of my electric car. I can get about 100 miles of range on a full charge, however I'd never plan a trip with my car of 100 miles even though my dashboard says that I have 100 miles of range. The lowest my battery gauge has ever been was about 8 miles left. Generally, when planning a trip, I'd never feel comfortable with having a round-trip distance that has less that 10 miles of wiggle room for my car's total range.
So, while my car theoretically has 100 miles of range, the real-world usage of my car that I feel comfortable means that I'll always have electricity in my battery that will never get used. In reality I'd just never feel comfortable while driving for a trip if I wasn't 100% sure I'd be able to make it home. That means ensuring that I will have that wiggle room in my battery charge.
I think of retirement planning as the same thing. The FIRE calculators are great for planning, but if you plan your retirement too liberal with regards to risk (lower success rates), then that means that you're forgoing having that wiggle room in your savings. The retirement calculators have an advantage that none of us will have in retirement...hindsight. None of us know when we will die. That means that I'd never feel comfortable having only a handful of years left in my portfolio in my later years even though I may only have a few years left to live. The reason being is because many of us will likely have no idea we only have a few years left to live.
What this all means is that with a success rate of 80%, if the goal is to have a feeling of a safe retirement amount, the real success rate is closer to 60-70% because many of those scenarios, while they "succeed" for the time frame you chose, they came drastically close to failing toward the end of the time period. Because I will likely not have any idea on when I will die, it will be very worrying for me to live on a savings portfolio that has a few years left to last.
You can certainly plan for that, however, by assuming that you'll live to your longest life expectancy. So if you're a healthy individual, you can plan your retirement as if you'll live until 90-100. That obviously has a impact on your success that gives you that wiggle room for later in life.
Whenever I see discussions about success rates, I always see discussion about speculation on how well the economy will do and disaster type planning, such as wars and economy collapses, etc. But what about positive benefits and breakthroughs that could also have a disastrous impact on your retirement such as medical breakthroughs? What if 20 years from now there is a breakthrough that cures a disease that you would've died from allowing you to live to 100 instead of 75. That's an extra 25 years that you hadn't planned on needing to account for. Things like that could be a drop in success rate of 10% or more.
Long story short, I wouldn't go any lower than 90% for myself, because the reality of a 90% success rate in historic calculators is more like a 80% success rate in real-life mainly because of the idea that none of us can predict the future and living on a savings portfolio for any of the scenarios that just made the "success" cut would be too risky for me to worry about later in my life at old age.