The new option caseInsensitive works great and it simplifies grammar a lot. But there are some lexer rule exceptions where it should be disabled. For instance, case-insensitivity affects lexing process for the following rule because a string starting with lower n is incorrect in TSql, MySql, but a string starting with upper N is correct. Enabled caseInsensitive allows both strings:

options { caseInsensitive=true; }
STRING: 'N'? '\'' (~'\'' | '\'\'')* '\'';

It's possible to check online the following string: SELECT N'sample';

I suggest adding a new lexer rule option caseInsensitive that naturally disables/enables case insensitivity for certain rules:

options { caseInsensitive=true; }
STRING options { caseInsensitive=false; } : 'N'? '\'' (~'\'' | '\'\'')* '\'';