Within the context of the article, the bigger form of the argument can be phrased as such:

• DirectX is not cross-platform
• OpenGL is cross-platform
• Blizzard is successful
• Blizzard releases cross-platform software
• It is more successful to release cross-platform software
• It is more successful to use OpenGL than DirectX

This is bad and wrong. As a snap judgement, it is likely that releasing cross-platform software is a more successful thing to do but using that snap judgement to build bigger arguments is dangerous.

This is an example of an appeal from authority and fallacy of division.

As it is, the heuristic is "X is successful and Y is part of X's business plan, so Y probably leads to success".

But Y doesn't lead to success. If I say, "Blizzard is successful and making video games is part of their business plan, so making video games probably leads to success," something should be obviously wrong. Why would it be true if I use "always releases Mac versions of their games simultaneously" instead of "makes video games"?

If you think their planning is no better than chance, or that Y usually only works when combined with other factors, then disagreeing with this heuristic makes sense. Otherwise, it seems like it should work most of the time.

As far as I can tell, the emphasized part is the whole reason you should be careful. Picking one part out of a business plan is stupid. If you know enough about the subject material to determine whether that part of the business plan is applicable to whatever you are doing, fair enough, but this is a judgement call above and beyond the statements given in this example.

Affirming the consequent, in general, is a good heuristic.

Maybe, but it is still a logical fallacy.