Just based on observation, the first 1000 rows are loaded, and then other rows are loaded as required.
My understanding is that 10 million rows is a “marketing” limit. In practice, you can have as many rows as you want.
I’ve not tried it, but I would expect that CSV export will work no matter how many rows there are.
A chart is just a different type of collection, so the same limitations apply around computed column types. Other than that, Big Tables work fine as the source of a chart.
We are in the process of putting Order history data into Big Tables.
That would mean 3 big tables
Orders
Items (OrderID included)
Bookings (ItemID included)
Its a bit challenging to understand the costs involved and the best design taking into consideration that user device performance is very very important for us as its our major problem using Glide for our operations team.