What you need to do is make a model in Hornresp. Model it with perfect sides, a smooth exponential horn.
Just for the sake of the argument make it 250 cm long, a throat area of 500 and a mouth of 6000. Put about 65 litres in the driver enclosure and drive it with a good 18 (BL 25, Mmms 200)
Now have a look at the impedance curve in half, quarter and one eight space and then look at it as an infinite horn. Surprise surprise – as the mouth size approaches that of an infinite horn the impedance goes flat, no harmonically related spikes.
Now try it as a collection of straight-sided conical sections – provided they approximate the horn curve there will be no additional spikes in the impedance curve, it will look exactly the same as the first “smooth” horn over the intended operating range.
You should now know where those spikes in the impedance curve come from and that for any practical horn there is not much you can do about it.
Bottom line is that the discontinuities in the thoat (excluding the mouth) will not cause a problem until the wave lengths become short in comparison to the size of the discontinuities. There is not much you can do about the mouth other than use a large horn or stack of horns …. and the LAB and many other horns do sound very good!
If you want to go higher then the discontinuities are a problem (short wave lengths). Have a look at Bill Fitzmaurices folded mid horn designs which use smooth curves to solve the problem.
www.billfitzmaurice.com