I adjusted the C2 adjust using the Zenith service manual’s guidelines so that the composite video signal at test point C1 is 8.1V DC when the AGC test point E is grounded and the tuner is set to an unused channel. The streaking persists.
So I used my cheap $50 handheld oscilloscope from AliExpress to view the composite signal at test point C1 for one horizontal line of the grid test pattern. It seems to show ringing distortion, but it may just be an artifact of my crappy handheld scope. I will confirm tomorrow using a real oscilloscope.
If the ringing is real, then what does it mean? Does it mean the only way to fix this is with an IF alignment?