According to the Jan Heine blog post I referenced, many high-end cranks don't pass the "racing" level of the test. There's different tiers.
Without testing all "classic" and "boutique" cranksets personally such a statement is really just supposition, since there's no database of passing cranks. Many OEMs request certificates of conformity from us stating that our components conform to the appropriate norm; some of these require only an affirmative statement while others require data sets and other info. Others either request no info or perform their own testing.
While there are different norms for different use cases, it's not up to the manufacturer to select a norm that is not reflective of the intended use of their product. Each norm contains rather specific language dictating the type of bike to which it pertains.
I haven't dealt with anyone that intentionally uses a less stringent norm not reflective of the intended use of their product (manufacturers are more open and cooperative than one might expect at ASTM, EN, ISO, or UCI meetings). To be fair, the people you encounter at these meetings are generally actually manufacturing their parts and perform other tests beyond the minimum in the EN.
Others I'm sure are not so diligent. There are many wheels that do not actually pass various aspects of the EN norm, just as when we were in the crankset business I had a crank from a rather prominent competitor last 463 cycles on the EN test.
Some countries in the EU are much more stringent than others about adherence to the appropriate norm. In France, for example, they will buy a bike at a shop and then subject all components to the appropriate norm. There are large fines and stop sale orders that accompany a failing result (so I hear, fortunately we have no direct experience with this).
Lastly, we should regard a passing statement for the pertinent EN standard as an assurance of quality of safety. These standards are currently under review as they were initially written when aluminum components were much more prevalent than carbon; as such, many of the failure criteria or tests as a whole are not ideal for current product.