When Teams content is collected from AeD and placed into threaded transcripts, I've noted there have been many challenges when you need to export it and review either natively or in a review platform. I've listed some of the issues I've come across and wonder if there are others in the community with similar views and how they have resolved them.
HTML output - PDF was removed and now is only available in HTML. This does not render well nor does the embedded metadata for a thread appear in the HTML output (ie. Channel name, thread ID, time zone, etc...). Embedded Images, gifs, and emojis do not render and are often treated as attachments thus making it difficult to identify. A PDF output is needed and there should be more content in the body to allow for native review.
Load File - unless you choose to export to PST (which does not export to PST), the output is placed into GUID-named HTML files with a load file that cannot be understood based on the field provided. Additionally, you cannot easily map these fields into a review platform because they are not recognized.
Join/Leave information is not available in these threads.
Output of the HTML files and attachments are in a flat output. There is no way to easily identify if a threaded export was part of a channel, sub-channel, or private channel within a Team. A hierarchy is needed to help identify this content. The file names need an improved naming convention as well so that they can be identified by TEAM_CHANNEL_DATE or similar means.
There are improvements being made all the time, but the general assumption is that AeD is the only tool that should be used for O365 discovery is a bit much. Organizations are complex and need options when on-prem and cloud data need to be viewed together or sent to external parties for processing and review.