Wilson is a good example, IMO, of a brand that had an awesome sounding speaker (the original X1 Grand SLAMM) which would work well even with low powered amps and made their speakers progressively less sensitive and arguably better measuring but, again IMO, throwing the baby out with the bathwater. I have a friend who owned the MKI version of the X1 and it sounded really alive and worked well with a mid-power SET or push pull triode amp. Later models (X2, XLF etc.) progressively lost this capability.No. That is ridiculous. No manufacturer decreases sensitivity on purpose (efficiency is not exactly the same, and there is no rating given to the public).
YOU may think it was not for an improvement in sound, but Wilson would disagree. They change drivers and crossover design to push the evolution of sound quality. Some thought older tweeters were harsh at different times but the new ones do not get that criticism. Guess what driver typically is easier to tame from ringing but at the cost of sensitivity? Tweeters without acoustical gain are a somewhat finicky device that basically determines every speakers sensitivity. All the other drivers can have limits but in other frequency range you can multiply the drivers to increase it or limit the bandwidth and adjust the driver. Tweeters on the other hand already only play a select region, and you cannot add more of them without causing horrific problems. The best you can do is have a line shape with no breaks in it (tiny braces do not affect it), without any sort of forced directivity. Sometimes the answer is very minor acoustic gain from a tiny amounts of directivity to get them to match other drivers to save some sensitivity - this is because a tweeter can have high sensitivity in a range you cannot cross to so it inherently has to be brought down if the mid cannot go high enough in hz to meet it.
There is no benefit to reduced sensitivity in itself. No one is making a speaker and just trying to reduce it for arbitrary reasons. In fact you cannot just insert a resistor without changing the shape of all the responses from all the drivers. It is extra work that is just unnecessary. They typically lose plenty to make a speaker work from the start and none of them want to be pigeon holed to a tiny subset of amplifiers.