How do those of you who study Taiwan deal with the northern standard that Skritter is focused on?

The standard response is to put a comment in the definition or mnemonic.
There are 3 basic classes of problems that I've run into so far:
1) Final tone is not 5, but follows tone of the individual character. Skritter accepts if I enter correctly for many words (though 窗戶 isn't one of them for some reason), but for this situation, a custom definition is OK because it pops up immediately after I enter the last tone. (But it's still a problem for pinyin practice because the definition usually doesn't pop up... usually definition practice is immediately after pinyin practice)

2) Tone of a non-final character doesn't match (i.e. 著急, pronounced zhao1ji2 in Taiwan). Thus, for a tone prompt of 著, I get it "wrong" and try to imprint in memory what Skritter tells me, and only find out that I was right (or that a different answer is right) after I complete tone of 2nd character. For pinyin practice, there is the same issue as noted in my previous item above.
There's also the issue that the sounds being repeated to me do not match what I should be learning. I've already had a word or 2 where I'm thinking of the "wrong" tone because I have the audio memory of Skritter's pronunciation.

3) Entire pronunciation of a word doesn't match. This hasn't come up for me yet, but obviously it will. This has all of the problems with #2, just worse.
Actually, this has come up for the audio pronunciation of a complete word (電影) where the Northern pronunciation of the 2nd character is more like yong2 than ying2, but only within the word.

Since it seems Skritter doesn't expect to handle other pronunciations in the near future (ever?), how do those of you studying in Taiwan make sure Skritter doesn't lead you astray on these issues?
(Except for these deviations, Skritter is the best way for me to practice tones, which I have been ignoring until I started here).

The related big problem for me is that I don't know WHEN I'm getting wrong information from Skritter. I only learned about 著急 because my wife was looking over my shoulder when that word came up. No clue if I have other problem words in my lists... (At least for situation #1 above I assume that the TW pronunciation is different).


For the Skritter Engine, I completely understand how early design decisions would make this difficult. I've spent plenty of time refactoring & restructuring my own code for new features/etc.

My preferred way to implement the option of Taiwan pronunciation would be to have a "dialect" selector. Then when pulling the pinyin/tone for a word, it's first checked against a dialect table, and then fall back to the standard table.
Obviously, constructing the Taiwan table would be a pain, but the "comments" in CC-Dict show many of these, and this might be easier to crowd-source than some of the previous attempts (though someone would still need to be a gate-keeper).

For problem 1), let us know in the study screen feedback when you see one of these words which can be made into a Taiwan-friendly neutral tone exception, and we'll add it to the list.

2) and 3) are unfortunate, but would require a lot of effort to solve relative to the size of the problem (which I agree is not a small problem).