So should the EPA be able to regulate noise pollution?
The problem with your definition is that it encompasses basically all of society. Everything humans do alters the environment, mostly for the worse. By embracing such an expansive definition of “pollution,” you’d turn the EPA into a super-regulator of vast swaths of the economy.
Software people should be able to understand the importance of separation of modules. You don’t put logic for process scheduling inside a file system module. It’s the same problem with treating CO2 as an air pollutant. Power generation is 1/6 of the economy. The mix of power generation sources the country should have isn’t just an environmental question. It cuts across society—it implicates the economy, national security, industrial policy, etc. Consider, for example, how Germany’s over-reliance on Russian gas turned into a national security problem for them after Russia invaded Ukraine. An environmental regulator isn’t at the right level to be making that decision.
Yes the EPA should be able to regulate noise pollution too, unless maybe if another agency is doing that job.
No, there is no issue to have that definition encompassing all of what in society causes damage to the open environment[1].
Note that your wording, "encompasses [...] all of society", does not correspond to my stance. It only encompasses what causes negative externalities in the way I defined.
It is crucial that there exist such an agency (EPA) that examines these environmental negative externalities (let's just call it pollution now, as it is so defined), and considers ways to limit it.
Each agency has its own axis, and it is the combination of them that is effective: if you only leave the energy sector to e.g. the USEA (US Energy Association), they won't have the incentive, definition, or analysis axis in the direction of looking at human health or forests decline, purely because it is not in their definition or their statuses.
Similarly, if you are left with just the EPA, they won't provide the expertise to evaluate if this or that type of electricity generation is feasible and with which kind of technology.
It just the division of expertise, the division of labor, and that's why scientists and engineers work together across domains!
[1]: damage should be defined relative to a severely detrimental impact to the biosphere or to human society. The environment is understood to be open as defined by not having clear boundaries that may, for example, keep the impact strictly contained within the limits of a volume property of the agent of this impact. Air is typically not contained in such a delimited volume, and is thus open environment. Water also flows freely and is connected to e.g. big bodies of water. And so on.
---
About "modules":
"Software people should be able to understand the importance of separation of modules."
The "energy" module cares about the watt-hours produced but doesn't care about the cancer caused by the smoke!! (or, for CO2, about the planet-scale extinction event caused by raising temperatures, or the multiplying forest fires)
Thus you need the "pollution assessment and mitigation" module!
There. Proved to you that I am a SE :))