The answer to this is long and nuanced. There is the complexity issue of both hardware and software and then the sheer amount of resource and capital it would take to thoroughly test something like this.
Software - If you've worked on any software product of sufficient complexity, you can have all the tests in the world but it's still difficult to think of all the corner cases and possible interactions. The best software companies out there release software with bugs all the time - you'd be surprised at how many issues show up in bug tracking systems. If you're a successful company, you can spend months doing QA, running automation + manually testing and your customers will walk the entire feature set in a day. Most companies have process to deal with auditing, hot fixes, critical bugs, etc. Even at the most diligent of companies, bugs get released into production. More than other car manufacturers, Tesla relies on software and probably has better software engineers and there are still this many problems. If you're a legacy manufacturer that doesn't know how to do software, you're either going to see the same or greater number of bugs or you won't be able to keep pace with the pace of software innovation in cars. If Apple can't release bug-free software, you can bet no car manufacturer can.
Hardware - Most legacy car companies are using the same mechanical systems (with the exception of the engine), derivative parts or systems that have had years of testing and refinement, and have a ton of experience with quality + manufacturing at scale. Anytime you introduce something new, there will be use cases and scenarios that weren't anticipated. Almost everything Tesla has done has been new and from the ground up where they had no idea how people were going to use it or what any of the unintended consequences would be. In fact, it's not at all unusual to see tons of issues anytime other car manufacturers try to do something new. The first year BMW introduced the 335i was a disaster - specifically excessive oil temperature and constant fuel pump issues. I know 3 or 4 people who bought that car and all of them had their fuel pump replace and were stranded at one point or another. One guy had to have it replaced twice, they also rebuilt his entire engine, his air conditioning stopped working, etc. That's without doing anything too new like FWD, brand new seats, giant windshield, auto presenting doors, etc. In fact, most new model introductions are incremental with mostly cosmetic refinements.
Cost - Testing hardware costs a f*ckton of money. You need to anticipate how people will use each system and build custom test fixtures to exercise and fatigue each part of the vehicle. This is imperfect because you can't simulate real world use perfectly. Imagine the cost to engineer and manufacture a test system for approaching the vehicle from different angles and having the doors auto present. Then you need to manufacture a significant sample size of vehicles that you're just going to junk to figure out where your manufacturing process has issues. In addition, you can spend all this money and the car looks great but over time a wire rubs against the body and creates a short - a problem that you would only find after a year of real-world driving.
It's an inexact science and a problem that's impossible to solve. It's a trade-off of cost vs quality and I think it would stress Ford and GM's resources to test something like this, not to mention a company with less resources and cash like Tesla. The car will never be perfect and problem free - just better and better over time. I'm convinced that the only way that a car like this gets built with quality is to have your customers walk find all the unknown unknowns for you and quickly addressing + fixing these issues over time. I'm also think Tesla agrees, which is why the Model 3 is going out to Tesla employees and reservations in CA first. It's inevitable that the early Model 3's will have issues and they want to keep those problems in the family.
