Modern post sorting offices have technology to read postcodes or zip codes etc rather than actual addresses so such errors should not occur. If I type the postcode for my actual house into Google Earth it zooms right in to where I live so how come the international post offices can't manage to do the same thing & read the postcode
This occurred some time ago so could well be before (reliable) machine readers.
I would also caution against implementing some kind of pan-European postcode system just for its own sake. The most important disadvantage is that I don't see that resulting improvements in postal accuracy (if any) would be in proportion to the huge cost and the unwieldiness that Peter mentions.
More specifically, postcodes are used for different things in different countries and have different levels of accuracy depending on the other preconditions for postal delivery in the country in question.
The UK postcode is extremely accurate, such that you can address a letter to personal name, house number/name and postcode and it will get there. It is also accurate enough to be used for a wide range of other things like car and home insurance. On the other hand, it needs to be that accurate, because the rest of the UK's postal address system is completely chaotic. Houses sometimes have names, sometimes numbers. House numbering is sometimes up one side of the street and down the other, sometimes odds on one side, evens on the other. 13 is sometimes missing. Chunks of numbering are sometimes missing because a bunch of houses were destroyed in WW2 and then replaced with something other than housing. Acacia Street, Acacia Crescent, Acacia Close and Acacia Avenue are all different roads and can often be located next to each other. (Compare the US, where the suffix (street, crescent etc.) is mostly irrelevant and it's the first part that matters.) There is no uniformly agreed way of displaying building numbers on the actual buildings. Counties/other administrative divisions can be pre-1974, 1974-1996, post-1996 or absent entirely.
The Swedish postcode (5 digits) is much less accurate by itself, but within Sweden there are no two postal districts with the same name, and within a given postal district there are no two streets with the same name (ignoring the "suffix" element).
Imposing a one-size-fits-all system would require significant (and therefore costly and confusing) changes to addresses throughout the EU.
Providing the destination country is clearly marked on the envelope, you don't need automatic sorting equipment that can distinguish a Swedish from a British postcode - it just needs to read the country. For international mail, the postcode doesn't need to be read until the letter reaches the destination country. It could
be useful to have an internationally accepted abbreviation for European countries analogous to that used by the USPS for US states, but only if it was recognised and implemented everywhere. Part of the problem with this is that there are so many subtly different one-, two- or three-letter country abbreviations in use for different purposes - vehicle number plates, languages, internet TLDs, postal services, sports competitions, you name it: they all use their own separate variant. So long as that's the case you'll get people using the "wrong" one on letters, and that's before you factor in the people who object to "UK" in relation to Northern Ireland or "France" for Corsica and so on.