Notice the wide range of the depth above or below the pump surface.

I know @jakeraby pointed this out long ago but is there an easy fix or just use new pumps?
These are cores that are getting reworked.

For me, I've decided I'd rather use a Type 4 pump.

The proper solution is to pin the shaft to the pump housing which involves very basic machine work.

So it sort of depends on the definition of what is an easy fix. Can it be done basic hand tools and a drill easily (<1hr) - no.

Can it be pinned easily on a vertical mill or by a machine shop in an hour or so. Yes.

In my case, the problem has been finding a single good pump. I've accumulated 5 to make two good ones and maybe a 3rd by doing some basic machine work that some would call blueprinting but I won't call it that since I don't have a print that is known that I'm working to.

I also intend to make a basic test fixture so that I can test & verify pump performance before it goes back into an engine.

I saw someone on classifieds looking for lots of cores to rebuild. I assume someone is looking to get in the business or rebuilding these. Not me, but, I say Hooray. Would be nice to have a viable option to find rebuilt pumps with a properly pinned shaft at a reasonable price below the insane cost of new parts and even then, the housing which is a key piece is NLA.

Photo in the top left corner looks like one I received, shaft way too deep. In my case, it turned out to be what I believe was a type 1 gear set in a Type 4 housing. Factory error or previous owner just making do? I don't know but it clearly didn't match any of my other pump shaft lengths.

How is the variation in shaft depth affecting pump performance? Are the gears still properly located regardless?

Shaft Position doesn’t affect gear alignment much until the shaft for the idler gear disengages from the housing.

At that point it goes from a double shear arrangement of the Type 4 pump (a good design) to a single shear arrangement typical of the Type 1 pump (less desirable IMHO). Single shear will allow for more deflection between the gears.

At its most extreme outward movement the shaft can begin to contact the cam bolts/rivets (a very bad thing). Pinning the shaft resolves the potential for movement.

"Outward" meaning "in" toward the engine? Doesn't seem like it can walk "out" toward the cover, it'll just stop there.

So where's the failure in this design? Inconsistent shaft length? It seems that if there's risk for the shaft to walk toward the cam, then that should have been a blind hole that the shaft sits in (with a hole for draining, if needed.

Pinning the shaft to the housing and letting the gear spin on it seems the "easy" button (it really isn't that difficult a machining operation).

There's at least one old thread on this but the shaft can migrate out far enough that it kisses the cam gear. Ouch!

Looking at the photo...which one is in the proper place?

Failure mode is overheating the engine opens the bore between the steel stationary shaft and the aluminum oil pump housing loosening the “grip” of the press fit. Oil pressure can then get behind the stationary shaft and the housing and “push” the shaft toward the cam gear hydraulically.

Agree completely that a blind hole would have prevented all of this. Blind hole cost more to machine and tolerances on shaft length would have to be tightly controlled. I suspect that is why VW did what they did.

Lower left, top center and top right look to be closest to where they should be.