You are guaranteed to be unlucky this time because the OpUp Gods only deliver manna from heaven when least expected.

The fare buckets are essentially a tool allowing airlines to work down a demand curve, segmenting the market.

When flights are forecast to be in high demand with plenty of high willingness to pay passengers clamouring for a seat, availability in the lower fare buckets may fall and eventually close altogether. When flights are not forecast to be in such demand, availability in lower fare buckets may be plentiful.

Note two points:
i) availability appears or disappears on the basis of forecast demand to come (net of forecast cancellations to come) and not actually the number of seats currently unsold
ii) flights may be full even if availability in lower fare buckets is plentiful if demand from high willingness to pay passengers is not expected.

BA probably nest most of their regular fare buckets (exceptions include redemption buckets).

This means that all seats available in lower fare buckets are normally available for sale in higher buckets. If there are 100 seats available in the Tourist cabin then Y bucket availability will be 100 (only 9 shown, see response to Fruitcake below) provided that Y provides the highest average fares. However there may be only 10 seats in the O bucket.

Now suppose 10 passengers buy seats under fares valid for booking into O class. Then Y availability drops to 100 - 10 = 90 and O closes out.

But now suppose that 10 passengers buy seats under fares valid for booking into Y class. Then O availability will either drop to zero or remain at 10, depending upon BA's optimisation process.

Furthermore, at a given point in the future (called a 'Data Collection Point' or DCP) BA's optimisation process may determine that the previous demand forecast was incorrect and re-optimise, either releasing or removing O inventory without bookings or cancellations.

Sometimes an oddity can arise. With class hierarchy Y>B>H it is possible to see availability as B>Y>H if the expected value of passengers booking B fares is higher than passengers booking Y fares. This can happen for example if passengers booking corporate deals with discounts buy a disproportionate number of the seats sold under fares valid for Y class inventory.

I believe that BA relies increasingly on other categories of segmentation tools, namely pricing. By this I mean that they segment the market by penalties for changes and cancellations, advance purchase and length of stay rather than closing inventory classes.

An airline relying on price controls rather than inventory management may find that the optimum solution is to leave all fare buckets open and let the pricing rules work their magic.

For the reason why they may often be the same, see my comment on price vs. inventory controls above.

The numbers never go beyond 9 because this would reveal commercially sensitive information to competitors and the travel trade.

For example, imagine a travel agent saying "you have 150 seats left on the flight, and I have 70 passengers wanting to buy, so you have to give me a big discount (meanwhile I will still charge the passengers the published fare)"!

This is an extremely important point. Aircraft offer more or less fixed capacity and airlines can upgrade if necessary.

So if there is no more demand to come and 5 seats left unsold after expected cancellations, all in F, then it makes sense to accept any passengers who materialise and reaccommodate as necessary.