A pretty good primer on basic speaker measurements is John Atkinson’s talk an RMAF about 10 years ago. He gets into sensitivity about 3 minutes in:
One thing that you’ll note is that out of all of the hundreds of speakers that he’s measured, the median/mean sensitivities are 88 dB and 85 dB. If you flip the boundary switch on an FR-30, you gain nearly 2 dB in the mid/treble from the 87 dB spec. Would I say that this is average sensitivity? Yes.
Do manufacturers overstate this number quite often, playing a bit of a specsmanship game? Yes! Some manufactures are worse than others regarding the accuracy of this spec, which seems to feed the confusion around it.
Is higher always better (as the OP suggests)? Why would a designer choose a higher/lower sensitivity for a given design? The laws of physics are pretty simple about all of this and there are simple design tradeoffs. Feel free to google “hoffman’s iron law”.
It states that the efficiency of any speaker system is directly proportional to the enclosure volume and the cube of the low frequency cutoff. Basically, you can have two out of three - small box, deep bass extension and high efficiency. To put some more dimensions to it, for the system to play and octave lower (20 hz vs 40 hz) at the same efficiency, the enclosure volume requirements go up by 2^3 for a factor of 8 times the size. Conversely, if you want the enclosure half the size for a given low frequency extension, the efficiency has to be reduced by half. We chose what we feel are the optimum balance of these three factors for this design. Other companies may differ in their approach and which of these factors they are prioritizing, but by John Atkinson’s presentation, you’ll find that most speakers are rather similar in sensitivity (with some notable outliers).
Note the careful use of sensitivity and efficiency. When speaking about sensitivity we are talking about the output with a fixed voltage. Most amplifiers are voltage sources and so will put out more power into lower impedances. However, this places larger current demands on the amplifier and performance of amplifiers often gets much worse on difficult/reactive loads.
Wilson audio was mentioned earlier in this thread as having rather high voltage sensitivity of around 91 dB/2.83V. They are great speakers but are often designed with a punishingly low impedance dropping to 1-2 ohms in the bass, with sometimes a significantly reactive load (looking at things like an EDPR rating, their effective impedance can be 1 ohm). For each halving of impedance, you gain 3 dB of voltage sensitivity and so you can see that the real efficiency of these speakers isn’t particularly high. Though the voltage sensitivity is rather high - the amplifier is called to deliver more watts in the form of current demands.
As Paul mentions, the FR-30 only varies from 3.5-5.5 ohm from 25 Hz and up) and is 5 ohm in the bass, making it easy to drive.
Sound pressure level in room is more complex than your calculator suggests. The room acoustics and seating distance play a significant role in this. Page 104 of Ear Geddes “premium home theater” has a pretty good description of direct vs. reverberant fields in listening rooms.
http://www.gedlee.com/downloads/HT/Home_theater.pdf
Basically, the more “live” the room and wider directivity the speaker, the more sound pressure is sustained in the room. Essentially, sound doesn’t actually fall of at 6 dB per doubling of distance.
Additionally, many high sensitivity speakers sacrifice not only bass extension in their design but also total bass output capability because the most significant physical parameter driving sensitivity on the woofer level is mass and the mass of the coil itself is a large factor on a high excursion driver. Essentially, you can cut the coil length and increase sensitivity of a woofer but the amount of excursion and bass output that the woofer has (and air that it can displace) can be dramatically lowered. Again, this is an engineering compromise that isn’t captured in a single number. Sensitivity is not equal to output capability.