It's important to compare dBs in a consistent manner. Things like distance, frequency Scale (A, C...), and meter response, proximity to buildings and so forth can make huge differences.
Went through a bunch of dB stuff on my E30 M3 recently. Got black flagged a few times for excessive loudness although no meter was in use. Usually the SCCA uses a 95 dB at 50 feet, A scale, slow response approach but they have no final consistent spec at this time, and local concerns trump this anyway. Site owners respond to letters and angry calls from locals and neighbors who don't use meters.
Some anecdotal information:
E30 M3 has an 8000 RPM 2.3 liter lump with high flow cat leading to Borla muffler. At one meter, revving in neutral peaks at 117dB (A scale: mostly measures sound over 1000 Hz, slow response). Since sound drops 6 dB every time you double distance, the reading would be in the low 90s at 50 feet, borderline.
I added a magnaflow resonator after the cats and before the Borla and the readings changed downward by 8-9 dB. No more black flags. Same basic sound but louder. Most of the removed sound was in the range the meter measures, leaving a good portion of the mean sounding lower frequencies intact The resonators are straight through and don't harm output.
I tried a few "at-the-track" mods such as fiberglass stuffing one tip, hose clamping perforated metal soda can in place, etc. Those cut sound but noticeably hurt output too. To get a large sound reduction from restriction as in the soda can approach requires something like an extra 8-9 PSI of backpressure - not good.
Changing the final muffler amongst aftermarket options changes the SPL by maybe 1-5 db: not that much in many cases. Adding another stage as in what I noted above seems to be more effective but is pricey and can't be done at an event.
Is the Elise a cat-muffler approach or is it cat-resonator-muffler?
It would be interesting to see the meter tests on the various Elise options.
Stan