American English speaker here.
1. I'd say that the non-phonetic approach allowed this to happen in the first place. In languages where the writing system was pinned to a phonetic system, there is no room for saying something incorrectly. There is room for replacing words and sounds but at least the writing reflects the spoken versions.
2. If you want to go that far, then we ought to just adopt Chinese characters which go a step further and abstract out pronunciation from meaning. That is, it is possible to understand groupings of Chinese characters without even knowing pronunciation. But this system also creates additional problems.
3. I think we have the technology to "bridge" our roots to whatever we create in the future. Today it is actually possible to make something live for eternity thanks to the internet. One good application would be English writing reform if we are to accept that it is the linqua franca.
Consider the possible benefits:
* 2nd language learners will have a much easier learning curve and thus be able to transition from their first language to English more quickly
* World communication would increase--I think this has war-stopping potential (or peace creating potential)
* Translations would hopefully become more accurate (less secondary translation/interpretation needed)--things written in "new" English would be more accessible to a larger audience.
* Areas dominated by native English speakers would open up to a larger audience, thus creating more competition and more contributions by people of other backgrounds other than pure English.
For example, right now we are seeing dominance in technology/internet by English speakers, but why should that remain the case? As such traveling to a non-English speaking country often feels like going back 5-10 years because much of their tech experience is based on companies paying for translations or waiting for local companies to copy the technology.
I actually think we have the capability of making "new English" for both existing native speakers and new speakers. For example for native speakers, we could write an input method that automatically identifies words, spelling, or grammatical structures that have been deprecated and may not be understood by "new English" learners.
We also have the ability to write and document English to new English translations and historical references. There will certainly be some fallout but we currently don't document many language shifts that happen even among smaller populations.
I'm sure many exceptional native English writers and culturally engrained speakers will balk at such an idea. My challenge would be to have them learn Chinese or some other not so closely related foreign language to the same fluency they can speak English in order for them to understand the pain of learning a new and complex language. Of course I am not happy with the loss of culture, but I think for the advancement of the world at large, such a path is inevitable.