Where to start... hopefully I can cover everything.
First of all, this has been beaten to death in the FI section.
A) Smell. Please read closely. There are two origins for the bad smell people notice. 1) The original 'test pipes' that street cars had were actually gutted cats. The gutted cats quickly deteriorated and gave off the same small as rotting cats do [very unpleasant]. People then deduced that test pipes will give off the same smell by incorrect conclusions. 2) On vehicles in which the test pipes were installed but no tuning was done, the a/f is often off from what it should be. The most prominent 'problem' is running rich. People smell the extra fuel and blame it on the test pipes. 99.9% of motorcycles do not have cats in any form, especially outside of California. Do they smell awful/odd when they go by?
B) Power gains [or lack there of]. Every car and every setup is different. Bottom line, test pipes do increase power. I could go on and on about this in great detail to counter all the misconceptions regarding back pressure and all sorts of nonsense. The best gains are when the piping before and after the test pipes are most effective. If you have a great header and a stock catback, test pipes are going to have a minimal effect. If you have a stock header and an 'ok' cat-back, they will have a little bit more substantial effect but still relatively minor. 5-7.7whp is what I'd expect from a properly sized test pipe setup with good parts on each end on a n/a car, with the right tuning. Test pipes are still often one of the best bang-for-the-buck aftermarket parts out there.
Knowing the fundamentals of how engines/power is produced in general, it's hard for me to understand anyone who says test pipes don't increase power. But don't take my word for it, go see how many race cars run cats to get faster lap times. It may take a while.
C) Sound Volume. Again, it depends on the particular setup and car and most importantly, how do you rate/communicate the difference in sound effectively? This is how I put it, for a given exhaust system, a test pipe will increase the total sound volume by 10% or less. However, that 10% maximum may contribute to significantly more 'raspiness'. That's impossible to predict accurately as just minute tuning adjustments or even the installation of a particular exhaust component can effect this between cars. In general, I'd say the added volume is quite minimal. When people say test pipes are "LOUD" it really makes me second guess what they consider loud and or want to hear their particular system before and after.
As for the 'octane', on a stock compression n/a car higher octane isn't going to do anything substantial even with "perfect" tuning. Higher octane is designed for power plants that were built in advance knowing higher than 91/93/94 octane is what the engine will run on best; usually 12:1 and higher cars regarding n/a applications. Of course there are exceptions but most "gains" from higher octane are usually on modded cars that weren't tuned correctly in the first place. (Small note: check out the LS460's stock compression for that 4.6ltr V8 that runs on 91!)
Lastly, I have never heard of anyone being fined in any state at any time for running test pipes. I've also only heard of 1 or 2 occurrences when the officer even looked under the car, both instances were to make sure they weren't running open header as I understand it.
