Paired by use (up-down-up-down) is better operationally where the slow and fast lines carry separate services which don't interact very much. This was true of the Midland for example when it was built (the slow lines were mostly goods lines and didn't even have platforms at many stations) but much less true today when Thameslink crosses over several times per hour, conflicting with the opposite-direction track in between.
The Midland is a particularly interesting case because it ended up with so much four-tracking and because very often the widening was not done all at once by adding to tracks on one side, but by adding first a track on one side, later a track on the other, and then later again converting the pairing by direction to a pairing by use. And because of the nature of the company's principal traffics it made very good sense to have the additional tracks normally as goods lines and to keep ordinary goods traffic clear of passenger traffic and fast goods (both of course using the main lines).
As edwin_m says, the coming of Thameslink changed things dramatically on the London to Bedford section. Instead of putting in some long loops to the slow lines, so that TL could keep to its own pair of tracks, the cheaper alternative of crossing (on the flat, of course) to the fast lines was taken, involving a conflict and all the joys of approach controlled signalling every time it happened. As levels of service both on TL and the Midland main line grew, problems duly emerged. I can think of a number of locations where the BR/MML on-time HST in which I was travelling was routinely slowed from line-speed to 20 to 30 mph whilst a slightly late-running TL was either got out of the way across to the slow or caused a conflicting movement crossing out path.
As for Slade Lane Jn, well, the PicVic scheme had a flyover planned and authorised for that location, and I certainly know that the idea was at least discussed again more much recently. (But this isn't Reading, or Airport Jn, or Crossrail, so it'll never get done.)