My understanding is as follows, referring to the data in your latest file:

1) You have a list of products with prices as per 1 August. I call this one "old".

2) You have a similar list as per 2 November. I call this one "new". It has item from "old" (but not all) with the same or updated prices and it has new items.

3) You need a combined list of "old" and ""new" containing the latest prices for all products that exist in both listings. I call this one "current".

4) Finally, you create a table called "PriceListTable" that links to various sources, on of which is "current".

The attached file contains a Sheet1 where I have "old", "new" and "current" side-by-side. The latter is created with PQ by appending "old" to "new" (i.e. not the same as "new" to "old") into a new query, with duplicates removed and sorted by product code. (In your real life situation you would connect both "old" and "new" to CSV files on your own system.)

My thought was that, next time, you clear-out "old", rename "new" to "old" and create a table "new" from the latest ARKMillsitems, and refresh the queries that will then create a new "current" table. Links to your PriceListTable should remain intact.

It's a far less complicated that setting up self referencing queries as I initially suggested.