Most microwaves emit, at least some, power in the 2.4GHz band. That’s where your wifi signals reside. Since the microwave will be between 800W and 1200W output, any leakage from any point is likely to significantly upset the wifi which, IIRC, transmits with a power of notably less than a watt. It’s about signal-to-noise ratio, if the microwave leaks even a fraction of a percent of its power, the signal from the router will be mashed. Doesn’t mean it’s unsafe. Cheaper microwaves do seem to do this, but I never seem to have a problem with ours (Though I’ve no idea if ours was expensive or not). It could be noise from the transformer/drive electronics that drive the magnetron that’s doing the wifi in, not necessarily even the microwave “signals” themselves.
Microwaves are remarkably good at what they do, very power efficient (almost all of the power put into them is converted to heat in the food) in comparison with something like an oven.
Microwaves don’t heat from the inside out, they heat roughly evenly throughout.
The microwaves are not tuned “to the frequency of water molecules” – that would be notably higher (approx 5 times higher), they are lower because higher frequency waves wouldn’t penetrate the food very far, heating the skin and not much else. It also heats loads of other molecules like fats and sugars.
As for boiling an egg with a phone… Assuming you could direct all of the phone energy in one direction, which you can’t easily, to the egg. You’d need approximately 15,000 joules to heat the egg to about 80 degrees average temp. At 2W (2J/s, the power limit for moblies) that would mean it would take approximately 2 hours to cook. But that’s all assuming no heat leaves the egg at all, through conduction/convection/radation. Which is virtually impossible too. So in answer to the egg question – no.
HTH