XTRF removes umlauts in memoQ project names

When creating memoQ projects in XTRF, we've noticed that it strips the project name of accents/umlauts, but only in the project name in the PM dashboard, not in the project name in the segment metadata.

For example, we type 190301: Kunden A Geschäftsbericht 2018 | Ergänzungen in XTRF, which shows up in the PM dashboard in memoQ as 190301 - Kunden A Geschaftsbericht 2018 _ Erganzungen, but is 190301 - Kunden A Geschäftsbericht 2018 _ Ergänzungen in the segment metadata field "Project". All three are different

This makes searching for individual project TMs in the master TM in memoQ almost impossible because the TM editor requires the full project name as it appears in the segment metadata for the filter to work. 

memoQ think that XTRF strips the accents/umlauts intentionally because memoQ itself is capable of handling accents in the project name. Is there any way to prevent this? (At least until memoQ provide keyword searching in the TM editor filter dialogue.)