I think the plan behind bip101 was that the average internet connection will likely keep getting faster with time, even today some people have 1Gbit fiber connection at home, others are probably watching 4k netflix right now.
Average internet connectivity improvements only give about 30% per year optimistically. BIP 101 was far more aggressive than that.
And while some minority of the world has 1 Gbit available, requiring it would be saying nobody else can use Bitcoin.
I'm only slightly rural and I can't get better than 5/0.5 Mbps yet.
A minority of the world has >1Mb/s persistent internet, persistent power and a computing device capable of running Bitcoin. If some people can't run Bitcoin full nodes now, it's not a failure - nor will it be the case if some can't run it in the future.
There are literally hundreds of millions of people that have persistent home broadband sufficient to run with larger blocks. It's not relevant that you aren't one of them.
The idea is not to run nodes only on fast Internet backbones. Remember the original idea was to have a truly distributed system. Scaling should be slow, slightly above the hardware improvement rate. We have plenty of Paypal wannabes and many altcoins where seemingly simple parameter change caused failures later on.
I agree with gmaxwell, changes should be carefully made, not to infringe on the original vision. If you want to make paypal coin, go ahead, just don't call it a bitcoin.
If core loses this 'battle', I think the original bitcoin experiment will be over. Thanks to shortsightedness of people like Gavin and Garzik.