It is done all the time...usually because individual mobile devs for commercial software are more likely to know iOS than Android, and very few know both. Vendors who are serious about being cross-platform maintain expertise on all platforms they support.
The reality is that most companies who have a product involving a mobile app don't have dedicated mobile devs on staff to do the work. They hire consultants...which introduces a whole level of economic and political complication because the consulting firm is often weaker on one platform than the other. AAPL locks in devs for iOS and OSX by requiring buying a licence to develop for the platform and sell through their app store, and having a proprietary tool chain that must be used to develop. Once a small firm has all these sunk costs in licences, software and training, it's difficult to make the commitment to maintain parallel capability on a different tech.
"AAPL users spend more on software" has been true in the past. But a much bigger factor in the marketplace success of a mobile app is: how big is your audience and how compelling is your product? There are thousands of iOS apps that never even paid back the licence fees needed to create them.
Also, the economics of having a business whose only revenue comes from selling a mobile app are very different from a business with a device that requires a mobile app to be useful. Examples of the latter are: smart watches, the Anova sous vide cooker, high-end quad-copter drones, portable avionics devices like the Garmin GDL-39, etc. Products like this know from the get-go that a user isn't going to buy a new mobile device simply for compatibility with their app, so they have a much stronger incentive to support both mobile OSs ab initio to enhance the acceptance of their main product in the marketplace.