Crowding definitely works for some traits, but it has its limits. If you're selecting for frost tolerance or disease resistance in small seedlings or young plants, why not.
Even for taste, in a cherry, you can cram 4 or a half dozen into the space for one plant, and usually get a sample from all of them.
In my environment, I was able to get plants to produce determinate vs indeterminate pattern in about a ? half gallon pot I think. Or a 1 gallon. But I was unable to do this in for example the beer cup volume of soil, where they just wouldn't even try to make more than one cluster of buds, for the most part.
Larger fruited tomatoes won't produce at all if they don't have the resources.
At some level of crowding, you will get plants that just go vegetative and don't set or grow any fruit, at least in my environment that's what I've seen. Shade is an issue for production and also for fruit quality in some conditions - you can end up with uneven ripening (K defects).
Maybe it's a valid goal to eliminate those that aren't the most competitive plants, and if so, more crowding will provide the selection event. For me personally, there's nothing more disappointing than a plant that doesn't produce any fruit to taste. Growing a load of vine instead of a load of fruit is contrary my objectives.
(Actually I have one now that I'm anxiously watching - the only PL in my F2 and the only one of them that isn't growing a bunch of fruit by now.
Too crowded in sharing a 15gal tub with a sibling. Already voted least likely to move ahead, and I don't even know what fruit colors or tastes will be. )
If the primary goal is to assess fruit qualities like shape, color, taste, and you crowd to the max, you may be losing out on a chunk of candidates, where they may not produce due to the crowding itself or lucked out on resources due to some other confounding factor.
And of course you can't assess production values without using a full space per plant.
Last year I tried using 2-3 gallon pots to assess traits other than production, and so far it's a reasonable compromise, where I get enough fruit from each to get a proper taste assessment. I really didn't assess for amount produced, so this year I have half of the plants in 5 gallon pots to check for that.
So I would advise to experiment with different levels of crowding and see what the results are and what actually works for you. I have heard of large fruited types being grown in small pots where they are fed liquid ferts every week or quite often. So there are various ways that might suit, depending on your growing space and environmental conditions.
I have heard another breeder who said he grew 200 F2's and tasted them all in order to select one special tasting fruit.
Maybe it's true?
But in my experience, there's a limit to how many similar tasting fruit you can process and still discriminate between them. It's not just me, I saw the same thing with my taste panel. At about tomato #4, they start to say they taste "the same".
Realistically, if you have a half dozen plants to taste you can do that in two sessions, and then compare the best ones in a third.
And then you repeat your test to make sure you didn't miss anything due to degrees of ripeness, the fruit being slightly more shaded, or some other confounder of individual fruit vs overall performance.
I don't think there's any way I could taste evaluate 100 let alone 200 F2 siblings in one season.
For me, if I grow ten or a dozen F2's and don't find anything I especially like, I will shelve it and move on to another cross.
Some parent combinations don't produce what you hoped or expected. Would you find that special plant by growing a hundred? Maybe. Or maybe not.
This is why I prefer to make lots of crosses. Anything new that I'm growing that's at all promising, I'll make some crosses. I'm not shackled to any one project, I can cast my net in a way that produces interesting food (and some sauce) and works for me and my space and evaluation limits. If we have a crappy weather year and there's a line that disappoints, it's the other lines that go forward next season.
If the flavor genetics in a cross is at all promising, you should find something of interest in a dozen F2 plants. Good taste can even be dominant, and most of your F2's tasty and staying that way through the generations, with a few worthless outliers to 'sauce' in a tasty line.
If the parents taste genes are really divergent you may get all very different tastes, and in that case, may be worth growing more F2s to sample the full range. But I would still split it between multiple years for my own space and pace. And if the tastes are really all over the place, they are probably not going to be stable at F2 either. So select what you like best, and see what happens in a bigger batch of F3 or later generations.
Back-crossing is another way to reduce the space requirement to stack up all the traits you want in one plant.
