Oh boy, here we go!
Yes because you don’t have a method of producing it. You hunch it. Wing it. Sorta kinda put things closer together because you superficially think they’re more similar to each other. I can’t be too hard on you here because it’s a misconception I’ve had myself. But nested hierarchies aren’t actually constructed on the basis of mere similarity.
They’re actually based on phylogenetic trees, which in turn make implicit assumptions about the mode of character evolution. Some algorithms produce trees by working towards the shortest possible combination of branches (a sort of “minimum evolution” type of algorithm), others use probabilities using a specified model of character evolution to infer which tree is most likely to have produced the data used (DNA/protein sequences, morphological characters). There are still others I don’t know how work, and some I don’t remember all that well.
Regardless, the evidence for evolution is in the fact that different sets of independent characters yield highly similar trees (or nesting clades if you will) when subjected to the same algorithm (meaning if you construct your tree based on engine and gearbox characteristics, you can compare it to a tree based on wheel and windshield characteristics for example). And the fact that the data has high levels of tree-like structure in it.
Your job is to show that using the same methods, designed objects like vehicles do the same. Meaning you pick a subset of characters of vehicles, and then derive a tree using a phylogenetic algorithm. Then you show that there is similarly high levels of tree-like structure in the data set as there is in biological data. And then further that using a different set of characters, a highly similar tree is produced. If you could do that, you would have shown something surprising, and then we might be able to begin to discuss whether nesting clades really are evidence for evolution and common descent.
But that requires you to actually do the hard work and compile lists of characters from lots and lots of real-world vehicle models, and then use phylogenetic algorithms on this data. Something which you have yet to even begin to do.
for instance: do you think that bicycle is more similar in genenral to a car than to other bicycle?
Which one in particular? Which car? Which other bicycle? And similarity is actually not the criterion of interest.
Organisms are grouped phylogenetically by trees inferred from their characters, which can be physiological, morphological, behavioral, or molecular, or all of them combined. And when I say phylogenetically I mean a tree is drawn based on some model of evolution or philosophical assumption, such as maximum parsimony, or maximum likelihood, and lots of other methods. None of which are “which ones are most similar?”.
The kind of tree spit out by the algorithm on the other end might end up with a result that looks like it grouped by putting more similar things together, but it actually didn’t. And some times superficial similarity can be misleading. You might think a Giraffe is more similar to a horse than it is to a Hippo, and in a superficial sense that could seem reasonable. But if you actually look at their character traits in more detail, you discover that in fact the Giraffe has more commonalities with the Hippo than with the horse, both physiologically and genetically.