I understand that prior to the expiry of the Telefunken PAL patent,
Sony Trinitron sets for the PAL market actually threw away the
chrominance signal on alternate scan lines, thus landing themselves
back in NTSC territory. Those sets had a tint control, and I know from
personal experience that they produced a perfectly satisfactory result
(I only learnt the other day why they had a tint control).

Depends on what you mean by "satisfactory". Passable, maybe.

When you discuss something at length, you become aware of those things you
thought you understood, but didn't. (Well, I do, anyway.)

I'd always read that one could construct a PAL receiver in such a way that
eliminated the need for a manual hue control. I never questioned this, but
now it makes little sense.

There are two reasons for having a manual hue control:

The user can adjust the color rendition to their personal (and usually
incorrect) taste. *
The user can correct for incorrect burst phase.

That seems to be "it". As we've seen, these errors can be corrected by
adjusting the hue control, whereas the other error -- differential phase
shift -- cannot be so-corrected, because the timing errors are not linear.

Here's where I get confused. The line-to-line polarity reversal ** causes
the differential phase errors to be equal and opposite, and thus cancel out
when added (at the cost of desaturation -- but that's another issue).

However... If the burst phase is wrong, then there is no cancellation of
errors, because there are no "errors" /in the signal itself/. (Right? (???))
Therefore, I don't see how line averaging can be used to eliminate the need
for a manual hue control.

If anyone knows of a reference with a non-tautological explanation, I'd
appreciate a pointer to it. Thanks.

* Left to their own devices, the average user generally sets the color for
greenish skin tones. I wonder if Vulcan viewers tended towards a pinkish
error.

** It's actually line-to-line+2, because the image is interlaced.