It's just an interactive version, but it still doesn't explain how you can ignore the checksum without first decoding the entire base58 string. On a few tests with random keys, using the technique shown in the 2nd link I posted, the checksum characters on average affect the last 6 characters of the wif, but I did find a key that affected the last 9 characters.
You can't ignore the checksum without first decoding it. By converting it to base58 from base256 the last digits will be different and with each key it varies so it is not possible to know the checksum without decoding the base58