Adjust an already existing changelog snippet for PR 7220

changelog.d/pr-7189.md

@@ -1,3 +1,3 @@
 ### 🏎 Performance
 
-- Reimplement `get_submodules_()` without `get_content_info()` for substantial performance boosts especially for large datasets with few subdatasets. Note: might result for slightly slower performance when many subdataset paths are provided (e.g. via glob) to the function explicitly. Originally proposed in [PR #6942](https://github.com/datalad/datalad/pull/6942) by [@mih](https://github.com/mih), fixing [#6940](https://github.com/datalad/datalad/issues/6940).  [PR #7189](https://github.com/datalad/datalad/pull/7189) (by [@adswa](https://github.com/adswa))
+- Reimplement `get_submodules_()` without `get_content_info()` for substantial performance boosts especially for large datasets with few subdatasets. Originally proposed in [PR #6942](https://github.com/datalad/datalad/pull/6942) by [@mih](https://github.com/mih), fixing [#6940](https://github.com/datalad/datalad/issues/6940).  [PR #7189](https://github.com/datalad/datalad/pull/7189) (by [@adswa](https://github.com/adswa)). Complemented with [PR #7220](https://github.com/datalad/datalad/pull/7220) (by [@yarikoptic](https://github.com/yarikoptic)) to avoid `O(N^2)` (instead of `O(N*log(N))` performance in some cases.