What's wrong with this idea?

Usually, when we read of the SPS scenario, there is a SPS and a
dedicated rectenna farm.

To me, such a scenario negates two of the advantages of GSO, namely
the wide footprint into which a SPS can beam power and the never
varying power output of the SPS.

To me these limations present three different scenarios for
increased efficiency. What are the flaws of logic or engineering
that I may be overlooking in the descriptions below?

1.) A SPS for a city is designed for below peak power rating, and
the neighboring SPS, which would be offset by perhaps a few
timezones, can supplement with additional power to meet the peak
power rating. Necessary premise: Peak power demand will ebb in the
city providing supplemental power while it is rising in the city in
need of the power. I also, realize that the microwaves coming in
from the SPS a few time zones away will be coming in at an acute
angle, and thus not be as concentrated. And this might also cause
havoc with issues like restricted airspace.

2.) A super-sized SPS, with multiple transmitter arrays, serving
multiple rectenna farms in a narrow longitudinal band. This could
provide power balancing between different utilities. Also by
extending this concept to a larger scale, could a Super-Size SPS
serve multiple utilities across the hemispherical line, thus
providing power to summer peak loads while the other hemispehre is
experiencing winter?

3.) A transfer of microwave power to and from multiple orbiting SPS
in order to take power where it is needed. This prevents designing
SPS capacity for peak power load. Thus, a SPS orbiting above a
longitude in the night, can beam a portion of its power across
space, perhaps relayed by other SPSs, to a SPS above a daylight
longitude. This SPS would receive the microwaves, phase them into
its own transmission and beam them down to the rectenna. What effect
would the microwaves have on the communication satellites KU band
signals? Such a lateral transfer of microwaves would cut a swath
through all of the uplink-downlink paths.

Thanks for your help in troubleshooting this idea.