If you planted a bunch of different tomatoes together, they could cross pollinate. But you wouldn't know it this year. The fruits would still grow true to the plants this year.

So say you had a Dwarf Wild Fred and a Speckled Roman- both plants would have their correct growth habit and produce their correct fruit this year. But if they crossed, you could end up with any combination of growth habit and fruit characteristics if you saved seed from this years fruits to start next year. And not all the seeds could have the same combo of traits, so you could end up with plants that have different growth characteristics and different variations in the fruit.

If you plant 2 different varieties, you might reasonably expect some sort of of that mix of parentage to show up in future seeds. If you plant many different varieties, you might not be able to figure out which parentage is what in future seed.

The cross pollinated ones you might enjoy might be a crapshoot. You could try bagging that plant to prevent crossing with anything else next year, or cloning the plant to grow out.

Well, chances of cross pollination are not that high. Depends on the varieties used, some have more exerted stigmas that are more easily found by foreign pollen. I would say typically chances are around 5%. If you want to make sure that doesn't happen just look at 'bagging blossoms' .

Hybrids of course, will give you something different no matter what, how different depends on the hybrid. Some of them give a high variance, but quite a lot of them have very little variety in subsequent generations.

What you describe is similar to how Joseph Lofthouse breeds landraces. He probably intentionally crosses some plants, though. I recommend that you study up on landraces and Joseph Lofthouse. He definitely selects for desired traits. The idea there is generally to get plants that are more adapted to a local climate and set of conditions. It's not the way the industry tends to breed things.

Yes, if you plant enough kinds of tomatoes, you're bound to get some crosses. I planted about 30 kinds in close proximity in 2015, and about 100 kinds in 2016 (and about 105 kinds in 2017). I've had a fair number of obvious crosses (but most have not been obviously cross-pollinated). I've probably had more crosses than average.

Tomatoes have perfect flowers; so, a flower can pollinate itself. Also, crosses don't happen as often with tomatoes as with some other kinds of plants. If you want a lot of crosses, you're better off with squash (of the same species), melons and other cucurbits, as well as peppers. Peppers have perfect flowers, too, but they cross easily anyway, if they're the same and/or the right species; okra is probably similar there. Joseph Lofthouse is working on developing tomatoes that cross-pollinate more easily.

Any tomato crosses you get will be potentially a little to a lot different every generation you grow them (especially if you keep letting them cross, and if you don't try to stabilize them).

F1 hybrids are just crossed plants. So, you'll have similar results with those as with your crossed plants (except you probably don't know what either of the parents are; so, you may get more surprises).