This is something I have thought about and is something I don't fully understand. I don't think I could function without the miter slot or T-Track in my table. I have a Woodpeckers Super Fence but it is by no means a cure all for every router table procedure. The T-Track is handy for attaching feather boards which I use religiously, on both top and sides. The miter slot is handy for keeping cuts at a true 90 degrees or other angles when needed using a miter gauge. My box joint jig relies on the miter gauge for its proper use. I also reference my fence location off my Miter Slot, in particular for jointing operations. The outfeed side of the fence is adjustable so I set it with a dial indicator attached to a section of Miter Track.
I knew when I posted this I was inviting questions but I was tired and hoped the "think about it" would save me some effort. Shoulda known better...
So let's think about it together: Miter gauges are used to hold a workpiece at a given angle and pass it through a cutter at that angle in a path parallel to the centerline of the cutter. That's why it's critically important to ensure your tablesaw blade is parallel to the miter slot, any deviation results in poor cuts or safety concerns. Less so for a bandsaw but wide blades can give you a problem. Just as critical if not more so for a disc sander, it'll throw your work back at you just as fast as a tablesaw if your gauge isn't traveling parallel to the disc.
But a router is different: there's no effective length to the cutter, it's essentially a single point of contact between the cutter and the work. That's why you can use a simple board for a fence, clamp one end and swing the other to get your desired depth of cut. Essentially there is no parallel to a rotating router bit, just a tangent intersecting its' circle of rotation.
While you could use a router table with a miter gauge to make square or angled cuts on the end of the work, that's almost always done by the tablesaw, radial arm saw or miter saw. Routers just aren't very efficient at cutoff functions, almost all shops have saws for this. Common practice is to dimension the work and then bring it to the router for further treatment, a rabbet, groove or profile for example.
Since the work is already square or cut at the desired angle, it's a simple matter to butt it to the fence and slide it through the cutter producing a cut parallel to the end of the work. For small work it's easy enough to cut a backer board at the same angle wide enough to ensure stability, hold the work against it and pass it through the cutter.
I'm sure I'm not telling you anything you don't know but there may be others reading this thread that don't have your experience, my apologies for the wordiness.
You have a good point with the box-joint jig, I've always done box joints on the tablesaw with a dado stack. The only fence I've used on the router table besides the clamped board is my current Incra TS/LS/Wonderfence setup, it makes box joints easy but I have to move the fence for each cut. Setting up a split fence for jointing as you described is easy enough but could just as easily be accomplished with a good straightedge and a feeler gauge. If you make your own featherboards you can make them long enough to clamp to the edge of the table.
Please explain to me how they weaken the table considering both are rigid sections of aluminum extrusion. If anything they would help maintain the rigidity of the table top in my view.
Just my thoughts on this often approached subject matter.
Now to the nitty-gritty.... This is gonna get a bit technical, bear with me.
I took a look at your uploads, you have a pretty typical tabletop: 2 pieces of MDF laminated top, bottom and sides with plastic laminate. Pretty much just like mine except I used BB ply. Google "MDF Manual", scroll down to pages 10-11 and start reading if you want reference material.
MDF is not of consistent density throughout its thickness, the outer faces are significantly denser than the center. It's pressed under heat so the end result ia a panel that's flat and pretty much stays that way under reasonable environmental conditions. It's a balanced system, strongest sections on the faces, weakest in the center.
Now think about what happens to that balance when you cut a groove in one side. You release tension in that side and it expands. The other side stays constant, one side expands, now you have a bow in what was previously a pretty flat surface. And to compound the problem you've sealed off all surface area with laminate except where you cut the groove. Those bare areas are now open to absorb moisture from the atmosphere and swell, getting larger. Larger on top, constant on bottom, now you have a hump.
While your track may add some strength in its length, that's totally dependent on the size of the attaching screws and their frequency. MDF is poor at screw retention so don't count on much help there. So the table crowns left-right at the groove but less so at the cutter. Now your cutter isn't square to the face of the work.
Front-to-back is another consideration, this effects depth of cut as well as squareness. So now your cut is not only out of square, it's no longer parallel to the face depending on where you apply pressure. If you cut the groove to an exact fit at low humidity, you've compounded the problem. The MDF absorbs moisture and swells, the aluminum track doesn't. The result is a bigger hump perpendicular to the cutter.
All of this movement may only be measured in a few thousandths of an inch and that may seem insignificant but the end result is inconsistency and unpredictability in the finished work that could take a significant amount of time and effort with scrapers, planes and sanders to correct. On the flip side, if the gods are with you and you hold your mouth just right everything could come out perfect. But that's a big if....
So there's the reasoning behind my brief response. You asked...