Because if you take away challenges you take away creativity.
No, you don't. You want to see the most creative thing I've seen done with Minecraft, ever?
That was done on the XBox version in Creative mode.
Furthermore I am waiting for people who feel that having multiple implementations of power is a challenge worthy of preservation to continue their crusade against unification and start a topic to complain about the Forge Ore and Liquid dictionaries. Surely if forcing the players to use divergent methods of power generation and transmission is considered a boon to creativity, so too would forcing players to use divergent forms of ores and liquids. Obviously that is absurd, as absurd as requiring multiple systems of power.
Finally, I have actually found it a larger challenge to stay constrained to one form of power. For the past several versions I have been able to generate up to 72MJ/t of power from a 36HP boiler. It wasn't until this version that I have even bothered to make a 36HP boiler because I had no need to do so. Power generation is exceptionally cheap and plentiful for 3 of the systems (MJ, EU, UE) and the other two (RP2, Factorization) are seen as niche at best. Meanwhile power demands have been exceedingly low by comparison.
What has changed recently that finally got me to try my first 36HP? Three things. First, I dropped anything which didn't interact with MJ. Second, I have added more mods which were using MJ. AE, MPS, MFR. Third, sensible cost for power transmission across distances (TE's Tesseract). The end result is that my main base now has more things which have a constant power draw, more things which spike power draw, and a great way to transmit that power at cost to remote locations. IE, I have a larger demand for MJ so my builds have to be robust enough to handle the demands placed upon it. That, too, is a challenge. One that I personally find more engaging than mixing in 3 unrelated and equally overpowered power generation schemes which all generate far more power than any one system can reasonably draw.
Now, there's an argument to be made that Power Converters shouldn't be lossless. That there shouldn't be a method of generating power through conversions itself. And frankly, you'll get no argument from me. See my description of Tesseracts above. But having a method of consolidating into a single model of power generation and transmission is a good thing as it can bring parity between demand and generation. Parity which can bring its own form of challenge to builds.