@jimscreechy It sounds like you need some schooling in basic audio electronics to have any form of valid contribution to this thread. Your advice is just plain wrong.
Here's a couple of articles that explain it and reinforce the facts. If you have an alternative view, please support it with similar evidence.
en.wikipedia.org
Learn about the most common -- and preventable -- form of distortion known as clipping. Discover the causes of clipping and how it can damage your speakers.
blog.teufelaudio.com
These articles directly contradict your statement that
"Clipping doesn't kill speakers. Clipping is far more likely to kill your amplifer than anything else"
I am struggling as to why you fail to understand the importance of the input signal, both in terms of content and level. I'll do my best to explain that, along with what an amplifier rating means.
TLDR: Clipping is what kills speakers quickest. Excessive input level needs to be controlled either electronically or by a competent operator. Simply putting in a smaller amplifier will not stop failures unless other methods are employed.
Putting in more speakers will help with overall dispersal but may not necessarily prevent speaker damage.
OK, so the facts.
Amplifier Power:
Amplifier power output is generally referred to as Watts. This is made up of 2 components - voltage and current. The output voltage is derived from the input signal - as it follows what the input voltage does, while the current will depend upon the impedance of the speaker or load being driven. Watts therefore, equals voltage x current, where current (in a simplified model) = voltage divided by resistance.
Stated Power Output
The RMS (Root Mean Square) of the power was hit upon many years ago as a simple method to illustrate the average maximum power of an amplifier and is usually defined as 0.707 of the peak undistorted signal. So to work out the RMS Power of an amplifier, we would input an undistorted signal at a known frequency, connect a known impedance load (Probably a simple resistance, as impedance is just a frequency dependant resistance) and monitor the output while increasing the amplifier gain until clipping just starts to occur. At that point, we would measure the voltage being produced and the current demanded by the load and calculate the peak power. Multiply this number by 0.707 and we have a reasonable approximation of the RMS power. Because this is not a true representation of what a musical waveform looks like, this figure can only ever be used as a guide. A 500W amplifier is not necessarily suited to a speaker rated at 500W, other factors need to be considered.
Clipping and speaker killing distortion:
Clipping happens in amplifiers for 2 reasons. The first that the input signal is driving the amplifier so hard (loud) that the output is attempting to exceed the power supply rails, so the signal is "squared" off when it hits this limit. The second is when the current demand exceeds that available by the power supply, so the voltage rails sag and clipping results. The first is most common in poorly controlled systems - such as referred to by the OP, the second where an underpowered amplifier is used.
Clipping is basically distortion, containing lots of high frequency harmonics. If you look at the image below, you can clearly see the difference between an undistorted signal and the severely distorted clipped signal. What is evident is that the clipped signal generates significantly more energy for the speaker to try to reproduce compared to the undistorted signal, leading to a greater likelihood of speaker failure - even though the peak power output remains virtually unchanged.
Speaker Damage
Speaker drivers - excluding any crossover have 4 main failure modes, many of which are interlinked:
1. Mechanical / Shock - Dropping the speaker causing damage to the basket and suspension. Generally, the voice coil will jam or rub against the magnet assembly, causing a distorted sound, even at low volumes. Common in hifi speakers if poorly handled during delivery or setup.
2. Short term high voltage failure - the input signal is so large that the coil will fail in the same way as a fuse. Common in tweeters when over driven due to clipping, where huge amounts of HF energy is generated due to high order harmonic distortion.
3. Long term heat failure - the input signal is large enough to cause a build up of heat, which eventually causes the voice coil to become damaged and to jam in the gap. This leads to the mechanical failure mentioned above. Common in bass units, where very high currents are called for. Can lead to fires in extreme cases!!
4. Over excursion of the cone, causing the voice coil to change from being part of a motor to being a simple coil as it leaves the safety of the magnetic field. This can lead to either mode 2 or 3 failure, depending upon the severity of the overload.
So, bearing all this in mind, our ideal amplifier would be one that would never clip or distort, and would never output more power than the speaker was designed to handle. This is achieved in practical terms by using an amplifier that won't run out of power, so has a higher rating than the speaker, but has electronics or a competent operator that limit the input signal so that it will never overdrive the amplifier or speaker. This is common practice for large scale PA systems and an electronic solution is normally built into active speakers and systems designed for unattended operation.
System design and distribution:
My experience is that many problems with sound systems are due to poor design regarding speaker distribution and dispersal. Covering large areas but still controlling spill and ensuring equal coverage is an art that takes years to perfect. The end requirement for the audio must also be considered. Is it simply to make announcements - so clear, band limited audio is required, to provide chest thumping music beats to incite excitement in the crowd - so more power, particularly in the low bass is needed, or is it, as is most common to try and do all these things on a small budget??
For simple, predominantly speech systems, set up on a temporary basis, grouping speakers as a point source makes perfect sense. It reduces unwanted echo, delay and poor intelligibility, but it will mean that the sound level between the back and front of the listening area will vary considerably. For music systems, distributed systems might be preferred, but the design needs to ensure that electronic delays are used to time align the audio throughout the area, so that the focus remains to the intended point. In a stadium, it's quite usual for the speakers to be nearer the crowd than the pitch, so dancers and anyone on the pitch needing to hear the system will probably complain it's too quiet while the the audience complain that it's blasting their ears off.
Therefore, a local monitoring system with additional speakers with separate control could be deployed pitch side, pointing at the performers. These will provide them with the sound level they require, while keeping levels for the crowd at reasonable levels.
Conclusion
Hopefully if you have read and understood this, you can contribute to the discussion. There's nothing simple about high power audio systems and making broad sweeping and factually incorrect statements helps nobody.
School over.