corrections for better l10n

I have added linebreaks to the stories, so we can reuse the strings from the summary. Otherwise we have the same string to translate alone, and as part of a bigger string.

With this branch changes, translators only translate the phrase once.

The changes can be seen at the bottom of the gitlabCI job 'check_l10n_freshness', that lists the new and stale strings this MR creates.