After sending the above post, there was something rattling around the back of my mind about doing two similar batches to compare yeasts. So, checking through my old notes I found it back in 2019.
This was to compare WLP775 with WLP002 which had a reputation for leaving some sweetness. The blend was 20% Beurre bosc pears, 50% Pomme de Neige apples, 10% Crimson Knight Crabs and the balance of about 20% Red Delicious to make up the numbers. Given the mix of apples, it looks as thought I was using up all the windfalls and leftovers.
Start of fermentation to bottling took about six weeks. At bottling and before adding any priming sugar my notes say... WLP002 slightly dry/sweet, WLP 775 Complex, touch of bitter, O.K. sweetness. Six months later, I note that I thought that the WLP002 was slightly sweet and "inoffensive". That might be because I had "made up the numbers" by using red delicious which I have found not to be much good for cider.
So once again it looks as though 20% pears is a good starting point. In recent times I have been using heat pasteurising to stop fermentation at the right level of sweetness and carbonation (bottle about 4 or 5 SG points above the pasteurisation SG. If you have the right setup, I understand that kegging is a good way to produce sweet carbonated cider, but I have found that mucking around with bottling and pasteurising at the right point can be a bit of fun as it involves a fair bit of tasting, yum, yum, (i.e. you are never sure of the result until you open a bottle!)