had a look at the datasheet for the TDA9115 regarding the E-W output.
I noticed that for the pincushion control IC outputs a specific waveform for each scanline...
then this is used to drive a few transistors and with power from the 12v supply... biases the Horizontal deflection timings to even out the scanline times. more voltage, the pic gets narrower...when zero, it's wide since the sides are caved in, this means either the wave coming out of the IC is flat or the 12v voltage is dead...
seems the 12 volt works AND a wave is being output. something is stealing our biasing voltage.
notice a cap, C727 is tied to ground after our biasing voltage has been produced and goes off to the horizontal deflection circuit...
I'll bet this is toast and shunting our bias voltage to ground. sure enough... no good. replaced it and it's good to go again.