Loss of shift smoothness (oil related, not mechanical).

[DaveInIreland]

Yes exactly this. It's 100% oil related. A clutch doesn't magically become correctly adjusted when you change the oil. That's assuming it was out of adjustment in the first place, which I keep on top of.

One of my previous bikes (klx250) used flat tappets instead of roller rocker's, and I had the exact same issues with the oil.
At 8000 miles I checked the valves, and found quite significant wear on top of the cam lobes. Almost like a small wave of metal where the load would come into contact with the valve tappet. So if anything I should maybe be changing it before there is clear friction in the shift lever, and assuming in the rest of the engine too. Though that's not very scientific.

My old bikes have always exhibited the same pattern - round about 2000 miles the shift quality takes a nose-dive. Nothing to do with clutches or suggestibility or personal fallibility or any other BS reason that can be thought up.
These are bikes with hellish high miles on them and many oil changes over the years, and it's repeatable time and again. In fact, several times over the course of that, I've been forgetful of exactly how many miles are on the current fill only to be reminded by a deterioration in shift feel and realise that, yep, 2K has elapsed (again). Time for an oil change.
The bikes in question are a pair of GS850s and an XJ900, all high mileage and all good runners. Their sump capacites are all around 3L, so it ties in with your observation of the smaller amount of oil in your sump pegging out sooner. In truth, the oil in smaller shared-sump engines takes more of a beating than in larger ones.

 

[HangFire]

I would not rule out the clutch adjustment as BusyLittleShop pointed out. Ever noticed that when on center stand and cold started, the rear wheel will start turning despite the transmission being in neutral? It will not do that when the engine is is hot. So it is quite obvious that a wet clutch operation is affected by oil.
You mentioned that the gear changes become harder, that would indicate to me that the clutch is not disengaging fully and dragging a bit. Fresh oil might be masking this and as it shears and friction modifiers break down, the gear change is affected.

If I were you, I would pop the clutch cover off, measure the plate movement and compare it to the spec in service manual. There is no reason chasing different oils IMO, since as you already pointed out, they all eventually start doing the same thing. I highly doubt an oil will fix your problem, it may only mask it for a bit longer than other oils.

He's talking trail bikes and you all are talking center stands? That's the second mention of center stands in this thread. Can it be any more irrelevant?

But that's the way it goes here in this sub-forum. One guy's experience here (not you, KrisZ) is nothing but the purest data, but everything you say is mere anecdote. Never mind that contributor can't even correctly articulate was science and the scientific method is.

 

[ZeeOSix]

He's talking trail bikes and you all are talking center stands? That's the second mention of center stands in this thread. Can it be any more irrelevant?

The center stand was only mentioned because that's how the rear wheel got off the ground in the examples used - so context was there. There is obviously still some "clutch drag" going on with the clutch disengaged and the rear wheel is off the ground, even if the clutch is adjusted correctly on some bikes (max plate separation possible). There is more clutch drag from the oil viscosity between the clutch plates of course when the oil is cold and thick vs when it's hot and thin. I don't think that makes any real difference in shift quality because my bikes seem to shift just fine regardless if the oil is cold or hot. Maybe some bikes are more sensitive to that. With a constant mesh transmission, you don't even need to disengage the clutch if the shift is done when there is no load on the transmission - that's how "quick shifters" work.