I seriously doubt that this problem will be looked at in RM7 because so much of the developers' time is being spent on completing the RM8 rewrite. But I hope the problem will be looked at in RM8.
The last complete rewrite of RM was RM4. At that time, RM4 included support for UNICODE (or at least the UTF-8 version of UNICODE - there are multiple versions of UNICODE and some of the versions support more alphabets than other versions) . RM's UNICODE support was released with great fanfare because it meant that RM4 and subsequent releases supported most of the characters required for most western European languages. UTF-8 doesn't support such things as Greek, Hebrew, Arabic, Russian, Chinese, Japanese, etc., but it's still major progress as compared to support only for English. For example, UTF-8 support means that RM supports letters such as ä, ö, ü, æ, ø, å, and ñ. (I'm sure I'm leaving out quite a few of the letters that UTF-8 supports that are not used in English). I'm not sure how UTF-8 handles the accents and cedillas that are used in French.
RM4 didn't just support UTF-8. It also supported sorting the additional letters in an interesting way. Namely, the additional letters were sorted based on how they look. For example, to a native English speaker an a and an ä and an å all look the same, so RM4 sorted them all the same. I think the problem is that most native English speakers don't think of ä as a letter in its own right. Rather, they think of an ä as an a to which an umlaut has been added. The treatment of an ä as an a for sorting purposes is an example of a good idea in theory that doesn't work out so well in practice. For example, it caused problems such as the search for Löffler not working correctly.
Support for multiple languages in a program such as RM is a very complicated problem and I don't pretend to have all the answers. I suspect that no one person anywhere in the world has all the answers. Therefore, the best solution for the RM developers might be to support user control of characters sets, and also user control of sorting and searching that involves character sets. In a really perfect world, the following might even be possible. RM would support the user's ability to set up the program properly for German. And that RM user could then export his or her work in such a way that other RM users could import the same character set support for German that the first user had already figured out.