That red clipping means the wave form has maxed out and is probably FLAT-LINING at the max level. So rather than having sound-wave oscillations that SHOULD HAVE BEEN THERE... the music is now FLAT because there is no "higher number" to put there.
It has NOTHING to do with whether you will blow your speaker or not, except for what volume you blast your music at. You will only hear distortion because instead of getting nice waves with distinguishable frequencies and tones, you will get FLAT-LINE or dampening of wave forms at that high level.
Think of music like a PHOTOGRAPH. Imagine you are taking a picture on a day with light sky and lots of white clouds. The RED-LINING is like you have turned up the brightness so high that the sky/clouds kind of blends into one WHITISH COLOR and you can't distinguish the blue/clouds/outlines/shapes anymore of the sky. It is just WHITE-WASHED.
Same goes for music at the high volume end. Often when you do AUDIO DYNAMIC COMPRESSION (like is done for FM radio transmission) it will take music and squeeze a wider dynamic range into a narrower one, effectively tightening the volume level of the sound. The problem with that is, some types of music (like Classical for one) have certain parts that are very quiet, and some that are very loud. Your ear can distinguish these and it makes sense for certain types of music.
Rock music and dance and other pop usually will just be loud all the time. That's why you can get away with both DYNAMIC COMPRESSION and essentially you only need to capture the "top of the wave" variations because that's the only thing changing.
For example, if you had values from 0 - 100 representing the audio data, imagine a rock sound like this:
87 89 76 89 88 93 94 100 87 98 95.... etc... notice how all levels are mostly above 75.
So rather than encoding the full numbers, you could essentially just subtract 75 from each of those numbers and encode a smaller range of numbers (from 0-25) and then just tell the system to "add 75" to the playback when it decodes it.
It takes less bits to encode 0-25 than 0-100 range.
The problem is when your music ACTUALLY HAS data in it (like Classical) that drops to much lower levels, like 34 35 76 80 98 54 24, etc.... Notice here that you can't really drop the bottom end as much because you have levels which go way lower, like 34.
RED-LINING basically is like DYNAMIC COMPRESSION but at 100% setting. It will take levels that may normally be say...
98 89 79 88 95 100 123 143 115 109 125 98 99 104 ... etc.... and it chops it to....
98 89 79 88 95 100 100 100 100 100 100 98 99 100.
Notice that the "100" is what we consider RED-LINE. So you have essentially LOST information in that section of the song. You get no resolution, no detail, no contrast, basically nothing out of that area with the string of 100's because of the RED-LINE, it was maxed.
You are better off subtracting 50 from every single part of that data string, and keeping the information. So for example, the string...
98 89 79 88 95 100 123 143 115 109 125 98 99 104 ... becomes....
48 39 29 38 45 50 73 63 65 59 75 48 49 54.
The second line just is the first line minus 50. It is still fitting within the data resolution of the container file format (0-100) but keeps all the information there. It may sound a bit more quiet, but then you can just crank the volume. Normalizing usually will take the BIGGEST VALUE (in the above string it is 73) and will bring that up to 100 (so it adds 27) and then it applies +27 to all other values so that no information is lost. It just sets the ceiling to the biggest number, and everything else below.
I hope that makes sense.
Bottom line is, you can't damage your speaker unless you actually turn the volume up on the source. If your source is properly encoded and NOT RED-LINED and you crank up the volume enough, you will STILL damage your speaker. Also, if your speaker is not rated to handle the WATTS that your amp puts out, you will damage the speaker. It has nothing to do with RED-LINE in an Audacity file. But you normally would NOT want to record anything or use anything that is RED-LINED because it means that likely it will sound BAD, Distorted or Otherwise not properly recorded.