It is not just that A groups together with B, it’s that different subsets of characters reproduce the same or highly similar nesting hierarchies consistently. You have no reason to expect this with vehicles, as one set of characters will produce one hierarchy, and another set of characters can yield an entirely different one. It is the consistency of the hierarchy across different, randomly picked, and independent sets of characters(of a significant number, not just a handful ).
If you compile a huge list of characters (the more the better, hundreds preferably) and construct a matrix for many many different vehicles(20, 50, 100 or even more different types and models, bikes, motorcycles, scooters, cars, vans, trucks, buses and so on from dusins of manufacturers and periods of time), and draw random samples from it and derive a hierarchy for each of them, then if you really are correct and vehicles also form an objective nested hierarchy (which you have still not shown, and continue to fail to argue for) you should reproduce the same(or highly similar) hierarchy from this. You should get consilience of independent phylogenies.
No, you can’t just insist mindlessly over and over again that “cars are more similar to other cars”. That does not show that you get a consistent nesting hierarchy across many independent characters, all that shows is you have some intuitive grasp of what superficial characters define what we mean by a car. As in a category with subjectively picked defining characteristics.
But what does the phylogeny of cars look like? We are not merely interested in knowing whether cars consistently form a distinct clade from bikes (which incidentally it might), but whether different characters will produce the same (or highly similar) clade of cars (that is to say, do the same, say, ten cars, A though J, out of this much larger set of vehicles, always form the same clade, say, (((A,B)(C,D))((E,F)G)((H,I)J))))?
Until such a time that you or some other creationist gets off their lazy rear-ends and does this work, there is just zero reason to believe such a result should obtain.
What we can say is that there is no reason to expect that designed objects will inadvertently yield a nested hierarchy, hence you can not consider a nested hierarchy as evidence for design. But evolution by branching descent with modification inexorably results in nested hierarchies, thus we expect them from common descent, and therefore they are evidence for common descent while not being evidence for design.