update TODO

TODO.md

@@ -7,8 +7,6 @@ General:
 * Currently using VRS 2.0a-based libraries. For lifting back to VRS 1.3, some basic post-processing should be fine (annoying but shouldn't be too trivial)
 * Without access to a production DynamoDB instance, Gene Normalizer will be quickest and easiest to set up via a PostgreSQL data backend. That, however, requires an extra dependency group (noted in README). We might want to make a `pg` dependency group here or just include it in core dependencies.
 * On that note, I've only done minimal testing of how possible it would be to drop the gene normalizer dependency entirely, but it'd be nice to get there.
-* Some of the singleton/factory stuff might be cleaner as `global`s
-* make `silent` a global
 
 Alignment:
 * Pretty sure this is mostly done.
@@ -19,7 +17,7 @@ Transcript selection:
 * Tests will need some extensive mocking (or cassettes?) for reliance on UTA and other external dependencies
 
 VRS mapping:
-* In general, this stuff is still pretty rough. Tests aren't passing yet.
+* In general, this stuff is still pretty rough. Not passing informal tests, and needs formal tests.
 * Finish double-checking the SeqRepo storage workaround
 * A fair amount of small questions about conditions written to handle specific scoresets/edge cases
-* More testing. Can be ready for CI by manually patching the SequenceStore class (or using SeqRepoRESTDataProxy and cassettes).
+* Can be ready for CI by manually patching the SequenceStore class (or using SeqRepoRESTDataProxy and cassettes).