You're welcome, Cass. Indeed, learning is continuous. Gathering wisdom may be more sporadic...
rrRobbo, you may regret asking questions with so much scope...
I didn't do the Blue Peter stuff (I don't send in my solutions either), but I used just 4 lines of another piece of paper to indicate the deconstruction of row 7 and reconstruction of row 8, the required contents in row 1, and the old column numbers. I think the sequence of solving makes a difference. Dealing with all the unrepeated letters of row 8 first, then the letters occurring twice. When I reached an ambiguity, I bypassed it and returned at the end. There was only one ambiguity.
Your point about the final grid not being all real words is well made. And it prevents me from joining all those who declare the puzzle 'amazing'. If a setter starts with the plan of re-arranging letters of one column in another column, and can make an initial grid with real words, then I'd say that is quite good. And in fairness the solving was enjoyable. But I'm not dazzled by the achievement. If the setter had somehow come up with a final grid of real words, I would certainly have joined the hordes offering platitudes.
As to pre-internet Listeners, what little I can remember, is that they usually didn't extend beyond a knowledge of writers and poets. In the 1990s I don't recall ever needing any other source than TCD, ODQ and perhaps a few other standard references such as Atlas, Dictionary of Music etc...