Short answer: because this will be more than just a wrapper

Long answer below

This is a recurring question, from long-time holders and newcomers alike.

There is a misunderstanding here and we have to confess we have not been performing very well at dispelling it.

Truth is we are not trying to merely create a GUI, as in a GUI wrapper. Fact is these simple wrappers already exist (Jojatekok's Windows GUI, jwinterm's lightWallet, others and we could probably use Bitcoin wallet too).

What we are building is more than a wrapper, it is a full-fledged scalable GUI software. And scalable is the hard part, not GUI - remember, scalability is one of the three pillars of Monero.

What is scalable?

• Quoting fluffypony "Merchants, exchanges, and automated systems need several things, and the database is just the first step. Without walletnotify / blocknotify functionality, and without an RPC layer that supports HTTPS and Simple Auth it is immensely painful for merchants to build systems, the paradigm shift is too large."
• Mnemonic seed for increased security (why).
• Low RAM footprint ("the DB", presently Symas Lightning MDB but later a lot others, since we built a database API).
• e-commerce prerequisites, notably for payment processors.
• Low bandwith footprint (a framework for SPV, like Electrum). The remote node provides a similar experience in that one doesn't have to download the whole blockchain, but this is where the similarities end and SPV would be lighter than a remote node.
• Much better understood (thus secure) and refactored code. This is behind-the-scene stuff and not sexy, but when it comes to securing your money, this matters. The Monero Protocol is sound, but we are not that comfortable with the source code we inherited in April (and improved since then). Note: we are pretty demanding on security, so when we say that "we are not that comfortable", this means the security is already high, just not high enough :)
• Gitian binaries
• Various mainstream users niceties, such as an address book, an easy way to do bookeeping (including easily report to tax offices, because Monero is not meant just for l33t haX0r, but for everyone, including legal persons like corporations, NGO, political parties...)
• Viewkey for optional transparency (see Riccardo Spagni's quote on The Three Pilllars of Monero)
• Have a solid foundation for easy improvements later on (like third-parties improvements). Because once the (scalable) GUI wallet will be there, we expect things to accelerate - and we rather be ready for the acceleration. Fasten your seatbelt, Dorothy.

Why is scalable important?

Because it is one of the three pillars :)

More seriously, because once the official GUI wallet will be there, visibility will increase a lot. Which means more users who know nothing about Monero and later not even about cryptocurrencies. These people won't give a second thought. One does only one first impression. It must be the right one. Hence the necessity for a wallet than can handle the load. Be scalable, not that much on a technical standpoint than on a, say, "political" standpoint.

We hope we answered your questions and, as always, don't hesitate to provide feedback. We'll do our best to make this even more clear if necessary.

Oh, one more thing: we do not have any release date for the GUI wallet, not even tentative. The only thing we can say is that the GUI wallet requires the database to be operational - and the database is progressing very well.

To conclude, some commented screenshots of the wallet some months ago, as well as a video.