fix: handle empty and multi-context results in evalLnot

[mmsqe]

to avoid handle ~ValidateTx(txn) <-> ValidateTx(txn) = FALSE differently, which are equivalent in TLC.

[will62794]

When making changes to the interpreter like this, I would expect there to be a concrete spec that demonstrates the bug, and that the fix addresses this. Is there such a spec? It's not clear to me from your changes. Same goes for #84.

Also, the convention is to test interpreter changes by adding a test spec in this directory, which then allows for utilization of the framework for conformance checking against TLC (see test list defined here). The Playwright tests are moreso meant for testing of basic UI functionality.