Of course, I wasn't saying one should solely rely on one power system. If you can work with multiple systems, then that's cool. You can even make your own system, if you want to. I still fail to see how mods allowing the use of multiple forms of power clouds the future of a single one. If it clouds one, wouldn't it stand to reason that it'd cloud the others, as well? The same could be said for BuildCraft, after all. Thaumcraft's automining requires a lot of research and, without proper precautions, can have rather nasty adverse effects on the local environment. Damn near every tech mod has some kind of ore processing; even PowerCraft, which has no method of pulverizing and doubling ores, can still process ores using nothing but its own mod blocks. Ore processing is something everyone nowadays does, it seems. There are, of course, exceptions like vanilla BuildCraft, but its becoming much more common.
You can also use those mods in concert with IC2 to add on to it and allow them to play off one another for your advantage. After all, if you want different nuclear reactors, you could just as easily go with Big Reactors instead of either IC2 or Atomic Science. DartCraft also does armour, and if memory serves, so does Thaumcraft. More and more, I'm seeing people playing with mod packs as opposed to singular mods, hence the growing trend of balancing mods against one another to ensure fair play, like EE3 and the upcoming DynEMC mechanic. This forum itself is dedicated to a family of mod packs, and therein lies the fun: finding ways for the mods to play off eachother's strengths and weaknesses to provide an all-around richer experience.
Being one of the most ubiquitous mods out there, it'd take something pretty damned serious to oust IC2 from the position it holds. As it stands, its still kicking just like it used to; it might need a new coat of paint, but it still works and more is being put into making it work better as opposed to providing more toys, and thus, more performance issues and bugs. Honestly, I laud the IC2 team for this approach, because I think its the proper approach when your product has come to a feature-complete state. Once you've gotten what you want in it, making it all work better should be a priority, instead of tacking on more and more crap to appease the masses' growing infatuation with the "latest and greatest." That's the trap Mojang fell into, and it annoys the hell outta me.