The lower bushing is expanded to put it on the fork tube, but then it is compressed slightly as it is inserted into the bore, so it can't come off the fork tube. It is not actually compressed. If it was it would bind. The ID of the outer fork tube is a hair larger than the bushing. Large enough to allow a slip fit, but NOT large enough to allow the bushing to expand and slip off the inner fork tube. Think of it this way, (I'm just throwing out numbers as I don't know actual dimensions) the bushing fits in a .030" recess in the inner fork tube and has a clearance fit of .005 with the outer fork tube. If the bushing is pressed out against the outer fork tube it still has .025 of the inner fork tube holding it in place; it can't physically go anywhere.
The middle bushing is installed while the fork tube is in, and is compressed initially to fit inside the bore, then expands slightly as it seats in the slightly larger diameter made for it. That's enough to let the lower slip inside the middle occasionally during removal, and jam as you mention.
This is sort of right. The same thing happens to the middle bushing that happens to the top (as correctly stated below). If you measure the upper and middle bushing ODs you will find that the middle has a slightly smaller OD than the upper. That's because the bore of the outer fork tube is like an upside down Christmas tree. The very top where the upper bushing seats is the largest diameter. The intermediate section between the two bushings is slightly smaller. The diameter is smaller than the seating area of the upper bushing but larger than the seating area of the middle bushing. This allows a small lip for the top bushing to seat against but will allow the middle bushing to freely slide down to where it seats. The area where the middle bushing seats is a slightly smaller diameter yet. This diameter is slightly smaller than the bushing OD so that when the bushing is driven/seated into place it compresses the bushing. As the bushing compresses that gap it it goes away. This makes the bushing behave as though it was a solid piece of metal that does not want to compress any more. This incompressability is what makes the friction fit between the bushing and the outer fork tube to hold it in place. One thing to note is that the ID of the fork tube is actually a couple thousandths smaller than the OD of the bushing in its "compressed" state, not just the relaxed state. This is where the press fit comes from. There is no "expanding" of the middle bushing until you hammer it out of the seating zone and it springs back to it's relaxed state. There is no lip retaining the middle bushing like the lower bushing on the inner fork tube; if there was you would NOT get the fork back apart without destroying the bushings at a minumum.
The upper bushing is compressed to fit inside the bore, and if you notice, the ends of the upper bushing have zero gap and takes significant effort to seat it. And remove it, thus causing the lower and middle to potentially jam even more.
Absolutely correct. The jamming in caused by the lower fork bushing wedging up under middle. This is where the clearance between the middle bushing and the diameter of the "intermediate" zone of the outer fork tube come into play. If the middle bushing is allowed to expand to much, and/or the lower edge was damaged during the slide hammering, the lower bushing can wedge underneath the middle bushing. How bad the fork tubes seize together is directly proportional to how much force is used to wedge the bushings together. This is why it's imperative to not beat the snot out of the fork when pulling it apart. The harder the impact, the more likely you are to wedge the bushings together.
The lower has the most slop, the middle a bit of slop and the upper has the least slop.
Not sure what this info is based on. There shouldn't be any more slop than what is required for variances in thermal expansion, flex and such. I can't say this is not the case, just wouldn't expect it for various reasons.
All ultimately ride on a film of oil however the lower has enough movement to cause the most wear as it moves the most (mostly fore/aft), and the teflon is worn off.
Linearly, all bushings see the exact same amount of movement. Lateral movement is a result of clearance between the fork tubes and as the clearance is increased the load bearing area of the bushing decreases. A smaller bearing area means higher pressure. More pressure means more friction and accelerated wear. For a better (exaggerated) picture think of a soup can put into a coffee can with both laid on their side. The only contact between the two is a line along the edge. In the forks most of the wear is a function of the lateral force the bushing is subjected to, the area that force is applied to, and the surface finish of the surface the bushing slides against. I don't know how Yamaha finishes the inside of the outer fork tube, but it would have to be hard plated (chrome/nikasil) as the aluminum is too soft and rough to handle the sliding friction. If the sliding surface of the inner fork in GEN II bikes is any rougher (microscopically) than the GEN I bikes that would cause the accelerated wear of the bushing. Perhaps the inner surface was not as finely honed because engineers figured the loading of the lower bushing would be less with a three bushing design and spec'ed the surface finish to be less to reduce cost. Or perhaps the ID is not plated at all. This of course is speculation but I could see it happening. I could not tell when I had my forks apart. In a ideal environment the surface finish of the inner fork tube SHOULD be equal to that of the inner fork tube. However, the bottom of the fork is where all the debris collects in the oil and this debris accelerates the wear of the lower bushing that is bathing in it in a similar way that a rougher surface would.. That's why it's important to perform frequent oil changes to keep the oil clean.
I agree the majority of bushings removed appear to have little wear on the middle and upper. The lower and middle are the most likely to be damaged by removal, but the upper (if you were really cheap) could be re-used.
From all that I have replaced, I go straight to the heat for the upper bushing before I start the moderate slide hammer process you correctly recommended. If I get jamming, most times I've been able to unjam. Only had one so far that required my special V-block method, which most likely was the scenario enountered by the originator of this thread.
Heat is a good idea. Expanding the outer tube reduces the press fit to allow the bushing to be removed easier.