I'll jump in here since this board is not frequented by a lot of professional car designers. I have consulted for a few car companies so I have some insight into the process at a few different places. Like anything, the process carries company to company. I'll also send this to a few car designer friends to see if they can add some better insights.
For the sake of simplicity, think of 3 different work streams happening:
- 1) product planning. This is where a lot of competitive benchmarking and market research happens. They also build the business case for the product down to the cost of an e-brake... which is why sometimes you see a really nice car oddly have an old floor style brake, or underwhelming tail lamps, or based on an old existing platform. All these costs are determined way ahead of time, and sometimes the marketplace shifts (a lot of times)
2) engineering. Here the packaging of the vehicle is also being determined way ahead of time. How far back the dash is compared to the axil, track width, wheelbase, height needed over the wheels for suspension travel, etc.
3) design. Interior and exterior design happening on an idealized notion of what the car should be based on the brand direction, competitive research, and user persona to select a theme down to DFM (design for manufacture) work of the a side of every part.
In a good process these three work streams are tied together at critical dates. Checking in with each other, influencing each other. When you read about Ralph Gillies working with the platform engineers on the wheelbase and track width of the Chrysler 300, or Peter Schreyer having a 20 year plan to improve design and perceived quality at Audi from the Audi 5000 to the a6 in the mid 2000's.
In a less ideal process these 3 work streams are happening in a silo like manner, each group doing what it thinks is best (usually only from the perspective of that group), and then mashing it together. This is what happens when you see a concept car end up looking almost nothing like the production car vs the original: Viper, TT, New Beatle, where the production car was almost the same.
Early on in the design process it can be very styling focused, with the hottest looking sketches (that might defy all reality) winning the internal competition to go to production as the chosen "theme". Once the theme is chosen it becomes a DFM process (design for manufacture) and all of those stunning lines and shapes have to be translated into metal, plastic, rubber, and glass. If a company has a more integrated process the designers will be ideating with the engineering package and product planning budget in mind (and hopefully have some influence there as well in terms of where to trade dollars around the car, and push and pull things slightly, remember, a car is actually a system of many products, from a window switch to a headlight). In the silo system the design has almost no chance of not getting butchered on its way to production. To possibly make it worse, sometimes the design team working up the theme is not even the design team doing the DFM, so imagine how invested they are in a faithful execution of that theme?
Now there is a 4th work stream, and that is HMI/UX/UI, which is playing an ever more important role and finally getting the respect it deserves. The car itself was a pioneer in human machine interfaces inventing entire systems for controlling a complex machine that we take for granted now. Got to a museum with early horseless garages and you can see how little was figured out or standardized in the early days. Ideally this 4th work stream is integrated in as well. Just look at the dominance of screens in most of the new Mercedes interior, or in the Faraday Future interior, this is likely going to continue.
Some auto executive groups can fall victim of the very human mistake of selecting great sketches of terrible designs. When I was at Nike I had a designer working for me that had this problem. He was always doing very hot sketches of what he should have known would be terrible designs. For example, having 4 panels of material coming together in a "X", very easy and dynamic to do in a sketch, but as a shoe it is always going to have a little "rat hole" where everything comes together, be harder (slower, more costly) to manufacture, and have higher reject rates (more costly again). The frustrating thing is this designer knew how shoes were made, we went to China together many times, but he insisted in doing these sketches that management could be tempted to get all excited about and the prototypes would come back as terrible, the exact terrible that anyone who knew how to really read the sketch could see.... now imagine in the automotive scenario where the team doing the hot theme sketches might not even have the production experience to know how something is made and you have a recipe for producing beautiful drawings of things that can't be made.
Conversely, at companies like Tesla, where they don't really make concept cars (when I spoke to the head of design there this was something they were very against, everything from their perspective had to be buildable, and they often brought the engineering resources down to glass makers into the studio to work with the team... argue about the designs all you want of course, but I respect the integrated philosophy).
This is a bit of a rambling response, but I hope it sheds some light on how different things can be. I'll send this to some friends in the industry so they can hopefully correct what is admittedly my outsiders view point.