The nightmarish die-offs of legend and song were probably the result of somebody doing serious trauma to their dsb without removing their animals first.
The DSB does a wonderful job of keeping it's toxic elements confined to the lowest layers, and detoxifying those elements before they can reach the water.
Not disrupting the DSB also means uprooting plants as little as possible, so if you're impulsive like me, your DSB will teach you a certain amount of patience. I do plan on moving some plants in my aquarium soon, but I am going try to space the job out: carefully move a couple plants today, a couple tomorrow, so the sand bed stays as intact as possible. DeeDeeK, as you know from reading this thread, prefers to simply cut the roots off and leave them in the substrate,
I might further add to the mention of nightmarish die-offs of legend and the DSB's doing a wonderful job of detoxifying those elements before they can reach the water by pointing out that if bubbles are forming in a region of substrate, that means the water in there is already saturated and will absorb no further gas - in other words, diffusion is limited enough to concentrate the gas near its source. The water is what carries the toxic gasses etc., not the bubbles which rise from the substrate.
In a really nice, loosey-goosey sand bed, the toxins can actually diffuse rapidly enough into the regions where nitrate-reducing and aerobic bacteria can eat it all up. If, in this case, the anoxic/strongly anaerobic zone is disrupted and exposed to the water column, things'll be fine. So, what we want is for hydrogen sulfide and other toxins to seep constantly out of the non-aerobic regions.
Myself, I am a bit cavalier about moving plants around but it is true, those with extensive root systems get cut off where stem joins root. Trying to re-bury roots is frankly a pain to do well and to just jam 'em all in together results in much root melting so you might as well cut them off (or just trim them if the plant can't just regrow its roots) if you don't want to carefully bury them which disrupts much substrate.
Short of releasing saturated toxic substances into the water column, disrupting the sand has definite disadvantages, especially when the disruption extends into more than one zone aerobic and anaerobic/anoxic activity. First of all, if oxygen is introduce into either non-aerobic zone, the bacteria either switch to oxygen-based metabolism and stop doing certain useful things like denitrifying, or the bacteria can simply die. Also, if the aerobic zone becomes temporarily anaerobic, nitrifying bacteria will die en masse. Then it takes days for the oxygen to either return or to be exhausted and bacteria to either switch back, recover, or recolonize, and for diffusion to return to normal.
If you're like me and blindly have faith that your substrate is not saturated with poison anywhere, you probably move your plants around at will, with the understanding that the more you disrupt the substrate, the less bioload your DSB will carry. Don't forget that the DSB can carry quite a high bioload, given time to adjust. Even taking out a good fraction of its capacity can be done with little or no consequences to its practical capacity unless you have the tank very close to the DSB's maximum capacity already.
I forgot whether I mentioned this idea yet in this thread so here goes: As plants grow, they absorb carbon in the form of CO2 and it is used as a building block in the complex organic molecules the plant synthesizes and builds its tissues from while the oxygen is released. In other words, the plant is warehousing carbon, as well as the minerals it absorbs, in all of itself. So, when the plant dies, that stocked-up carbon is released, largely as CO2! The minerals likewise are released.
Buried in the substrate, the roots will decay slowly and CO2 and minerals will steadily diffuse into the sand. That CO2 will enter the water column and boost dissolved CO2 levels. I haven't measured the difference it makes, in terms of ppm non-DSB vs DSB but would like to. A friend measured my tank and said it was 9ppm, short of the 15ppm levels used in Dutch tanks buy enough to seriously enhance plant growth. I plan on buying my own CO2 kit and figuring it all out. Also, it's time to study the use of CO2 in planted aquaria (study in a planted aquarium?)(ha ha ha).
So, to attempt to explain myself more simply and clearly, if the plants are growing in a tank which only gets its CO2 from the atmosphere, they absorb it from the water and the air replenishes it. Duh. But now the tank contains the CO2 from the air and the carbon stored as plant tissue. So, the potential CO2 grows and grows until removal of the plants(including root system) by the aquarist counterbalances it. Leaving root systems in the sand allows the potential CO2 to grow until the aquarist decides to start removing them and a new equilibrium is reached.
As the severed roots decay, the rate of decay regulates the release of CO2. If new root systems are left behind to decay at the same rate that carbon is released from the substrate, the result will be a steady, consistent level of CO2 in the water and a steady balance of potential CO2 stored in the sand as the various carbon compounds roots are made of.
I favor Cabomba Caroliniana and Eichhornia Diversifolia (variable leaf water hyacinth) as attractive plants which develop extensive root systems and tolerate being severed at the roots and replanted very well. There are probably tons of others which are equally good. I just stick the stems back in right in the middle of the root system, next to the little stump. They grow very very well that way.