The HP 740B I own is another to keep an eye out for but getting a bag of rocking horse poo is more likely especially if you want one with the input and output boxes. These are 2ppm while the Fluke 332/5D are 1PPM.
I think you're off by an order of magnitude there. The Fluke 332/5D is specified as 10 ppm basic accuracy over 60 days.
Getting one from the USA and having to ship it via courier to avoid the black hole that is Postnord's customs handling makes it prohibitively expensive.
Does the eBay global shipping program ship to Sweden? For all it faults, it moves the burden of the customs paperwork and fees to eBay, so that may help in your case.
Right now there are a couple of good looking 5100B's from Poland on the German bay, the sad story is they are ca. 2.2K EUR each .
2.2k EUR for a 5100B is ridiculous in my opinion. Waiting for one to pop up on US eBay for $1k and shipping that should be cheaper. And even $1k is expensive in my opinion. That's similar to a Fluke 5440B, which, though it is less versatile (only DCV, no ACV, current, resistance), is in a different league in accuracy and linearity.
From this list in the first post, the Fluke 335D has the best specs, but it is a boat anchor and from what I've read somewhat of a bear to maintain. The older ones have a mechanical chopper. Later versions had a FET chopper.
Here is an alternative circuit that should perform better.
Some lighter calibrators are the Valhalla 2701B/C (quite close in specs, but from what I've read not as stable in real-life) and Data Precision / Analogic 8200. Shipping these from the US can be quite affordable. Don't get fooled by "buy it now" prices. If you're patient, you should be able to find much lower prices. For example, I see a Data Precision / Analogic 8200 that recently sold for $300, and also a Valhalla 2701B for $300. That is a similar to what the Fluke 335D usually sells for, but shipping should be much cheaper.
The Datron 4000A is much more modern than the analog Fluke calibrators, has better specs and can be fitted with optional resistance and current functions. Not sure how repairable they are.