Hello,
Welcome to the Community.
So you are looking solely at conversions based on Adobe data correct? You aren't trying to compare Adobe data to Facebook pixel data (since you said that you aren't adding pixel tracking to your site).
When it comes to Bit.ly links, in all the testing I have done, Bit.ly forwards the referrer through on clickthroughs... so the use of those link shorteners shouldn't impact your tracking... however, how the links work from Facebook can definitely have an impact...
In most cases the referrer from FB should be set... but you need to make sure that there is no rel="nereferrer" attribute added to your anchor tags... this is a directive that tells the browser to not set a referrer in the request headers, and when that happens, those links will be treated like Typed/Bookmarked traffic.
Also, Facebook mobile apps should but may not code a referrer onto your links. I have seen the FB app use m.facebook.com as the referrer... but that also might be coming from the normal FB posted links, and maybe not Bit.ly links... The only way to know for sure what is happening is to test...
Have you ever done proxy testing for mobile apps? You can use the same setup to test clicks coming to your website from the FB apps; this process isn't limited to testing your own mobile apps. This might be a place to start, set up Charles or Fiddler on your computer, connect your device through the proxy connection then open the FB app and test your links.. check the "r" value being passed to Adobe when your site page view is tracked.
This is really the only way to know for sure how your posts are responding... It very well may be that the apps are being treated as direct traffic...
The other thing you should check based on your known FB traffic, what is the average PVs/Referral... in our experience (and this may be because we are a News organization), but most people from Social Media were basically one and done traffic...
Another thing, do you have a mobile app? And does that mobile app use Universal Linking? Just in case you are unfamiliar, Universal Linking is basically, when a user has a mobile app installed and try to open a webpage for the site, there is an automatic detection that occurs that instead of opening the webpage, the app is opened to the equivalent page in the app....
Now, this can be a problem for your attribution because Mobile Apps don't track referrers at all! I actually posted an idea about this, with so many apps moving to Universal Linking, we really need to have referral tracking added to mobile apps....
Now in the above scenario, Mobile Apps don't even classify under "Typed/Bookmarked".. they literally have no referral data at all (no Referrer Instance), but this could very well be swallowing up your FB traffic, making Direct much more prevalent, since so many people use mobile for social media.
Now, as for your Navigation scenarios... assuming FB still does this, I really need to review this... but FB used to provide a modified User Agent String that you could use to detect in-app webviews.
You can see the Skill Exchange event where I presented this here: https://experienceleague.adobe.com/docs/events/the-skill-exchange-recordings/analytics/sept2021/anal... (start time 22:50)
Unknown is a bit of a mystery without seeing more in your implementation... I've not seen Unknown in my data... if data is Typed/Bookmarked then the Referring Domain should also be that... It's almost as though the Referring Domain is being stripped out during processing? (This can also happen if a non-url value is being passed into the Referrer.. Adobe will remove data that doesn't match expectations)
