I have seen inter-subspecies crosses. They look intermediate between the parents. For example, a cross between crookneck and zucchini looks like a large crookneck with an extra fat neck, or like a smallish zucchini with a somewhat restricted neck. Still perfectly edible as a summer squash.
Landraces have more complicated genetics than simple Mendelian inheritance. For example, when I cross a pink banana squash with a blue Hubbard, the first generation of offspring are football shaped and have orange skin. The next generation, there are some banana shapes, some Hubbard shapes, and lots of football shapes. Perhaps some round fruits will show up. However, the colors are now all jumbled up. There will be pink bananas, green bananas, blue bananas, striped bananas, orange bananas, etc. Also lots of different colors on the Hubbard shaped and the football shaped. If continuously grown as a landrace, the jumbled up shapes and colors will continue indefinitely, and some parental types will continue. I could reselect for any particular color, or any particular shape, or I could let the traits remain variable. I am most interested in flavor, and productivity. I cull very small squash, and very large squash. I want something that my cooks can easily handle, and that fits well into their cookware.
The middle squash is a hybrid: Shown with it's parental varieties.
Later generations:
Edit to add that there are 4 alleles with major effect on color in the original cross. 2 sets of genes. One set affects if color is orange/green. The other set affects if it is pale/intense. So that basically means that there are 9 fruit colors available from a single cross, some of which may only be available in a landrace with mixed genetics, and not in pure strains. And then there are genes for stripes or mottling.