I was wondering how do car modelling artists always keep a good topology on their models. Do they model it from one mesh to cover the entire car body or they model the parts separately with different objects? I was struggling to model a 1970 Chevrolet El Camino SS recently because some areas of the car like the headlights and door panels are difficult to model.
Do as you like. In the end the purpose will dictate you the object choice… for example tires as different objects in a game engine…
Most car modelling artists make 3d models of modern cars like Audi R8, Nissan GT-R, and many others. I am modelling a Chevrolet El Camino SS whose model year is 1970 and that year is very old when compared to 2021. Most car modelling tutorials are specific to one car only. Also, the way how you should model one car is different from another one because of how the topology will flow on that certain car. The El Camino has a pair of round headlights on each side of the front view of the car which was common on older cars before.
This is how the El Camino looks like on the front:
both! 
So I always model everything as separate parts because a) it’s more realistic, and b) it’s generally easier to get the shapes right on smaller pieces, and c) it’s easier for texturing.
However, shrinkwrapping is my recommendation. If you don’t know about shrinkwrapping you should definitely research it. For example: https://www.raymairlot.co.uk/blog/the-shrinkwrap-modifier-a-hard-surface-modellers-best-friend
That basically means you want to model the entire body as one shape without details, and then you will model the different pieces and shrinkwrap them to the general body model to ensure a smooth transition between the pieces. If you try to model separate pieces without shrinkwrapping you will never get the pieces to flow nicely at the seams, and cutting holes for stuff like headlights, door handles, grilles, etc., will give you topological nightmares and weird shading.
I’m currently (though extremely slowly) modelling a 1997 Camaro, and I’m doing it this way. I modeled the body as one big smooth shape and will model individual panels and shrink them to follow the body.
in fact now that I’m looking at it, I probably shouldn’t have even cut out the wheel wells yet because that is distorting the reflections.
I hope this makes sense. If not, I can try to model a panel or door or something to show how I shrink it onto this.
Ok, here’s a really quick example, so the panel edges look bad, but these are 3 different meshes shrinkwrapped to my body mesh. You can see they match perfectly to the surface, despite having holes for headlights etc. The reflection across the edges are continuous.
Thanks for the advice