Alan Mulally has an outstanding track record from Ford Motor and Boeing , but most likely wouldn’t be the right guy to lead Microsoft  to success.

Rather, what Microsoft needs to become relevant in the consumer space again are Marissa Mayer and Yahoo!

The new frontiers in the great ecosystems war are shifting very quickly. Google, Apple, Microsoft, Amazon and Facebook  are playing an ever-higher-stakes game across an ever-wider battlefield of frontiers.

In this battle of the giants — Microsoft, Apple, Google, Facebook and Amazon — it is not a matter of optimizing and fine-tuning an assembly line of steel and plastic-based boxes. To be sure, devices are still made (mostly in Asia) and sold through various distribution networks.

However, these American “Fab Five” technology giants are not winning or losing based on the relative efficiency of this-or-that assembly line. BlackBerry (BBRY) and Palm had access to perfectly good assembly lines — and look at where that got them.

Apple didn’t win market share across various product lines in the last five to 15 years because it had superior factories. Rather, Apple’s gains were about far more fundamental architectural choices in combination with product definition and design.

The difficulty facing Microsoft now is all about architecture and specific product definition. The architectural issue is about seeing the emergence of a whole new set of technological forces, and how they will enable new use cases. This is the level of problem Tesla (TSLA) is solving in betting 100% on the electric car, and that Google and Salesforce (CRM) solved in terms of creating cloud services to replace PC software.

The second problem — specific product definition — is one of taste-making. It’s not too different from the fashion industry. Would Ralph Lauren (RL) have become Ralph Lauren without Ralph Lauren? There is a reason fashion houses almost always wear the founder’s name — it’s all about one person’s taste-making.

Do you remember Frank Lloyd Wright, or do you remember the construction company that poured the cement on time? Exactly.

Steve Jobs was the taste-maker of the technology industry. From the Mac, to the iMac, to the MacBook, to the iPod, iPhone and iPad — along with much of the software — Steve Jobs all but had his name on Apple’s door.

There were no consumer studies, no rule by majority, no cowing to the conventional wisdom. It was only about one thing: Steve Jobs designed the products Steve Jobs wanted to use. You could join him or stay away. As taste-maker, Steve Jobs garnered more followers than anyone else. Apple became the technology world’s fashion house.

For Microsoft to have a chance at surviving — let alone prospering — in the epic battle against Google, Apple, Amazon and Facebook, it needs a new leader who has both an architectural vision, as well someone who is a supreme taste-maker who can make Microsoft into the technology world’s new fashion house.

I’m sorry, but this person is not Alan Mulally. Perhaps Alan Mulally would be a great VP of supply chain management at Microsoft — but not CEO.

Maybe there is no sufficiently strong taste-maker with architectural vision available to save Microsoft from its otherwise inevitable demise. But let me suggest that the person closest to fitting the bill looks to be Yahoo! CEO Marissa Mayer.

As I explained above, there are really only two qualifications that matter in terms of finding a CEO who can save Microsoft from what otherwise looks like certain defeat against Google and Apple in particular:

1. Architectural vision. Microsoft has been busy re-labeling itself as a cloud services company, and to be sure it is moving in the right direction. The problem is that its revenue and profits remain mostly in the old world of localized software and traditional server models. Even if Microsoft succeeds in its technological transformation to the cloud, how could it protect its profits, let alone grow them?

It is mostly Google — but also Amazon and Salesforce — who exposed Microsoft’s architectural weakness. In order to stem these losses, you might think that it would take someone from those companies to fully recognize the radical threat posed by the shift to a cloud architecture.

And know what to do about it.

2. Taste-making. In designing fashion, you need someone who (a) imposes a new form of beauty, while (b) avoids making stupid design mistakes. You have to have a good view of market trends, and see where there is an unfulfilled opportunity somewhere on the chessboard.

Then, you have to have the self-confidence and power to impose this will on the company’s development team so it can work with you to realize your vision as soon as possible. As General MacArthur said, behind every lost battle are two words: too late.

Can you see anyone but Marissa Mayer who fulfills either of these two criteria for Microsoft?

But wait, there is more!

Marissa Mayer is CEO of Yahoo!. It may seem like ancient history, but Microsoft did try to acquire Yahoo!. In an act of certified insanity by the previous management and board, Yahoo! rebuffed. It was one of the costlier acts of corporate self-delusion in history.

Given the rise of Google, Facebook and even Twitter, it seems more urgent than ever to ensure that Microsoft can battle these players even on this newly expanded field of social interaction and commerce. In order to do so, Yahoo! of all companies comes the closest to having the remaining pieces of the puzzle Microsoft needs.

Conclusion: Not only is Marissa Mayer the right person to re-invent Microsoft, but Yahoo! is as important for Microsoft’s competitive matrix as it was when it bid for the company last time.

There is just more thing: Enterprise versus consumer.

Increasingly, enterprise technology is becoming more different than consumer technology. IBM realized this a decade ago, and sold its PC business to Lenovo. Dell  got rid of its smartphone business. Cisco shut down its consumer camera business, and divested Linksys. BlackBerry is now getting out of consumer in order to focus on enterprise.

What it takes to run Microsoft’s large enterprise business may be materially different than what it takes to run its consumer business. On the enterprise side, the interaction with the customers is very different. The marketing and PR is different. It’s a different type of customer. If you’re running for President, you have to decide if you’re campaigning in front of the general population or if you’re campaigning in front of the economics club at University of Chicago.

This is to say that perhaps Microsoft has become too big, and it is almost impossible to have a new CEO optimize for both large enterprise and consumer markets. It might be necessary for Microsoft to divide into two companies.

If so, Marissa Mayer may be better suited for the consumer half of the company.

In either case, Alan Mulally isn’t the ideal candidate for Microsoft, even though he may have been the ideal candidate for companies such as Ford and Boeing. On the other hand, if Marissa Mayer didn’t exist, Microsoft would have to invent a Marissa Mayer in order to reinvent Microsoft’s consumer business into something that can compete with Google, Apple, Amazon and Facebook.

[Cross-posted at The Street]