If Microsoft had spent the money needed to actually increase VFP sales, that money spent would have needed to have been many times more than the profits from VFP sales, thus a business loss. The reason is .NET (and VB before it) was competition (and strategic for MS) against VFP. Microsoft didn't really market VFP since the late 1990s when VFP 6.0 was released. VFP was then killed internally at MS, but then it was decided to continue the product decreasing the cost of production and marketing strategy for each version, mainly focusing on keeping VFP developers to move to VB (and then .NET), or at least not moving off of Windows.

VFP 7, 8, and 9, was to only keep MS developers (and their runtime apps) on Windows, with about half moving to .NET and using SQL Server over time. VFP 9 actually was making a profit, but... the amount MS could make by killing VFP and moving the key team members to work on .NET would make MS more profit long term. MS focuses more on sales and profit, often at the sacrifice of loyalty and customer satisfaction.

Look at the massive number of products, services, and technologies MS killed or gave away over the past year alone (probably more than ever), including IronPython and IronRuby for the .NET platform. MS will lure developers to invest and bank on a strategy, but if it ends up not profitable for MS, they will pull the plug without hesitation. That's not to say many .NET developers are very happy with .NET and succeed well in using it. But just know that the primary developer platform strategy at Microsoft is to sell licenses of Windows, Office, Server products, and now also Azure/cloud.

>From your post:
>
>But by the time VFP 9.0 was released, the amount of sales for all versions of VFP combined annually was less revenue than Microsoft sales of Visual Studio in only one day
>
>Even though lack of marketing contributed to that, it says it all.