Singer was great about keeping records (mostly) and, even with the 'missing' information for some of its earlier models, it puts the other manufacturers to shame. W&W were not so good. Then, of course, Singer renumbered Bridgeport machines and effectively went back to 0 but with a 'W' in front.
So, there is a huge grey area around the D–9 and 9w. No-one really knows just when (in the sequence of serial numbers) Singer took over. Nor is anyone sure exactly sure when they dropped the Wheeler & Wilson decals and introduced the Singer Celtic Knot style. And later, when they changed the bobbin to a 221 style (with holes in the side) from the one shaped like a donut or bagel, depending on your cultural heritage.
So far, I would estimate yours to be a little before 1911, but not before 1909. So far, it is the latest example I have that uses the donut shape bobbin. If yours has the longer, Singer standard bed length, then it is clear that the company had adapted the smaller W&W beds to fit its own cabinets/treadles before it changed the bobbin.
There is every reason to believe that yours is a 9w7 but, in later parts books, it says the 9w7 took the 221 (later) bobbin. This is an anomaly I have yet to explain.
There is no record of how big a production run would have been at W&W, but we know how many of its popular models Singer commissioned - 100,000 per batch was not uncommon. Your machine is only 4,000 away from the batches which had the 221-style bobbin, so it could have been rushed out for a particular trade show, or promotion, as pre-launch samples or some other imposed deadline… with the rest of the batch following with the new bobbin. This is, of course, imagined but I know things like that have been reported. They had to have machines to demonstrate at the launch, so they could take orders for the main 'batch'.
Once they were tooled up to make the new bobbin and maybe its housing, it was a simple job to get the assembly line to fit these 3 or 4 parts rather than those 3 or 4 parts, as all the rest of the machine was the same. I think what I mean is: even though changing to a new bobbin sounds a big deal, practically, it was quite a simple change.
I have never heard why they changed bobbins. There seems no good reason for it. One is not inferior/superior to the other. Perhaps the later style was just more convenient - more in the style of the other bobbins they made. Maybe the old W&W bobbin-making machine was wearing out and they didn't want to replace it just for this one model. Maybe it was a 'size' thing - and they thought the user would prefer slightly larger bobbins, less frequent windings, etc. It could be anything. If you were able to post a request for serials - D–9 or 9w - I'd be very grateful. I can't do it myself as I'm stuck with a deadline and won't have time to monitor the replies regularly (I put together the ISMACS News magazine for them and it always takes more time than I have).