Lazily evaluate list items by treating lists as iterators

Lists are data types that exist at-rest and in-memory during points of execution in the graph. They are analogous to a `Vec` is Rust.

The idea here suggested by @TrueDoctor, if it turns out to be feasible, is that we can possibly use fields (lambdas, executing upstream nodes multiple times with different context data). This approach would be analogous to an iterator in Rust.

So in our analogy, instead of `.collect()`ing into a `Vec`, we could use the iterator directly. From a UI perspective, we can still collect into a list to display in the spreadsheet view based on where the user clicks a thumbnail to view the data at that point. But during regular rendering, collects don't usually occur and we therefore work not with lists but repeated field (lambda node) calls.

This concept requires further exploration by @TrueDoctor before it may or may not become a plan to be implemented.