Second, I'm using the online editor, not the downloadable tool. I've no clue if that makes a difference.
Ignore white space works pretty well, but doesn't ignore carriage returns and some other white-space-like control characters. It would be nice if there was a way to ignore all control or non-printing characters (see https://en.wikipedia.org/wiki/Control_character)
It would also be nice if the ignore white space or ignore non-printing characters feature would ignore them when showing other differences too. An example will explain better: http://www.mergely.com/sxiiFJkJ/
With the current ignore white space selected, I see that the LHS line is struck out and both the RHS lines are new (not helpful). I expected to see that the words "1. list" from LHS and "1.list" from RHS were the same while the word "test" would be struck out from the first line on LHS and the RHS would have added a second line with the word "test". If an "ignore control characters" setting was added and turned on, I'd expect the LHS and RHS to have no differences.
Why would this feature help? I compare lists frequently looking for every word and punctuation difference. Word differences matter a lot. Punctuation differences might matter, it depends. But white space, new lines, pagination, column breaks, they rarely mean anything - and if they do mean something, you wouldn't turn on ignore white space.
Minor caveat: would I want something like "Find the cat a tonic dude!" and "Find the catatonic dude!" to be marked as differences? Yes - they mean different things. But should that difference be identified when ignoring white space? No - nothing is perfect.