Tomato characteristics shifting by seed selection

I know most tomatoes come true to type - if you save seed from a particular tomato, unless you have one with promisciuous parts like Joseph Lofthouse, they will grow new plants that are likely to be the same as the parent plant. Apparently however minor variations can occur. I'm not sure the mechanism, but a local grower to me has been selecting a particular purple beefsteak for flavour and now has one that is apparently nicer than the typical one of that variety.
I could imagine that local bacteria might cause your plants to become used to your soil, but can only assume minor mutations over generations has make some difference in this case, as the plants were grown elsewhere in Scotland for the taste trials.

I haven't grown tomatoes for a few years, but have two plum/paste tomatoes this year: 'Amish paste' on the right which grew rather poorly but did make a few tomatoes, and an unknown 'fleshy red plum' which had much more healthy plants. The Amish paste does seem to have more flesh than the other plum variety, but presumably I could select seed from plants that did slightly better or fruits that have more flesh in proportion, and hope to make small improvements in time too.

Really interesting question. I've noticed my saved Brandywine seeds seem to be getting slightly better adapted to my plot over the years, earlier ripening, handling wet summers better. Never quite knew if I was imagining it or if something was actually shifting.