First off very few people are going to write memorable posts about how reliable their car has been. Humans generally have a negative bias (can't afford to get eaten by a saber toothed tiger!) and we tend to tune in negative things more than positive. Some people's biases are more negative than others, but the majority of humans are a bit on the negative side.
So we remember the people who had problems more than those who said they have had none, or only minor problems.
The reliability of the 2012 and 2013 cars was poor. Tesla had a lot of problems with those cars, but they were a brand new car design from a company that had never built a car from the ground up before. Problems were inevitable. The 2014s were better and the 2015s were better still.
The problem is nobody has any really accurate data about any make of car and reliability. All the information collected by Consumer Reports, JD Power, and anyone else collecting data is all volunteered. Even if they get a statistically valid sample size, the sample is usually skewed by those who bothered to fill out the form, and humans have a negative bias so probably more forms will have negative reports than positive compared to the real data set that only the car companies have.
Another thing that is going to skew results is customer expectations. Someone spending $20K on a car is not going to have the expectations someone who spent $100K might have. I have to say I've been surprised at the pickiness of some forum members about quality. I want the car to be right, but I can live with a few minor imperfections. I've been driving the same $22K Buick since 1992. It has some quirks that were there from the start that are still there. I still consider it a wonderful car that has been rock solid reliable. The only time I was actually stranded with it was when I got stuck in mud (I pulled over to the side of the road and the right front wheel dropped into a spring I didn't know was there), which was not a reliability problem.
When I picked up my car last Monday, there was some red paint mixed in with the white on the bottom of the frunk lid in the front. I pointed it out to them and the service center people seemed more concerned about it than I was. They took the car into the shop for about an hour and they apologized for not getting it all, but when I looked at it, they got 90% of it. They said they would get the rest next time I brought in the car (for service if nothing else). I'm fine with that. I expect some people would have had a cow about it.
Tesla knows exactly how reliability has improved, just as GM knows how reliable their cars are and Toyota knows their cars. Nobody shares that data with the rest of the world, so everyone tries to get data from end consumers.
Another thing that skews the reliability ratings against Tesla is the way the reliability ratings are scored. Reliability ratings all have weighted values in their scoring system and those weights were created for ICE vehicles. Tesla got lots of huge black marks on the scoring because they replaced a lot of drive units. By the scoring scales, that's equivalent to replacing the engine and transmission in an ICE. Pulling the engine OR transmission on an ICE is a major repair and only done under the most severe circumstances. With a Tesla pulling the drive unit is easier than replacing a spark plug on an ICE so Tesla decided to replace a lot of drive units that were functional, but not working completely as designed. In most cases because of unusual noises, but they were still functional.
As Samuel Clemons said, "there are three types of lies: white lies, damn lies, and statistics." There are people out there who hate Tesla who skew the statistics to make Tesla look worse than they are.
All new cars have lower reliability. Consumer Reports covered that in their Car Talk YouTube video last fall. Even brands known for reliability take a quality hit when they introduce a new car. The more redesigned the car, the bigger the reliability hit the first couple of years. CR even pointed out one of the tricks Toyota uses to keep their nameplate reliability scores up is to get more years out of each design and only change things slowly. Fewer changes means less reliability hit.
The more innovative the car, the more reliability problems it's going to see initially. The Model S had it's teething troubles and the X is having them now. I suspect the falcon wing door problems are going to plague Tesla for some time. The middle row seats will likely have issues for a while too. Tesla has been aggressive about pushing the envelope in the past and they have taken the reliability hit for it. It looks like they learned that lesson on the Model 3 and it will be a much more conservative design, though it will have some problems initially.
I've only had my car 4 days, but took my first mini-road trip on Friday (just to Olympia, WA and back). It behaved well on the trip, even in some really nasty rain and late afternoon glare on the way back.