Why all translations have spacing between some strings removed? It makes translating harder.

[K0-RR]

Why aren't there empty lines like in the en-us.json? Or should I ask why are there blank lines in en-us?

[jaruba]

Missing strings are resolved with a script (/scripts/add-all.sh). When the JSON format is (programmatically) parsed, the empty lines are lost. This a tech limitation, unfortunately it cannot be fixed.

[K0-RR]

May I then suggest removing empty lines from the en-us?
It would allow translating by opening both the en-us and the in-translation version in any text editor that supports side-by-side displaying (the line number would always be the same for any given key in both files)

[ngocanhtve]

Use Excel or Google sheet to separate the section to be translated into a separate column, duplicate that column, translate the entire duplicated column with Google Translate. So you have the translation and the original in 2 columns next to each other, convenient for correcting the translation for smooth in case Google translates graffiti.
Then use a bit of regex to merge the translated column back into the json file and commit it to the source.