Recently, a player posted a question on the developer's blog (well, more like an article), basically complaining that the game doesn't look as good as modern AAA titles. SerB replied this:

"It's simple. When you work on a project, you have certain resources. When it comes to the available money and work, you can spend them on graphics, or on the network part. We spent most of it on the network part - the money on acquisitions (purchase of the BigWorld license) and the work on polishing the networking code. Practically all online games have similiar priorities, that's why all online projects are behind with the graphics. If the project is successful, you can allocate your resources to graphics as a second step (with online projects it's the render), which is being done.
That's why you can't just "make a crossbreed between BigWorld and Crysis", because there are format differences in object handling, both in the network and in the client parts. Therefore, it's usually better to polish the already existing solution, which - again - is being done.

A player then asked, why doesn't WG simply make a MMO with the best possible graphics and then focus on the networking part?

SerB answers: "That can be done, but it's very expensive (and the worst part is: not expensive as in money, but in manhours of qualified employees). It's so expensive that noone found it worth doing for now.

The player then suggested giving the "graphics polishing" to an external developer, so it doesn't cost the employee manhours, just money.

SerB answers: "That would be possible, but it would take a long time, as the developer would have to understand the part of the game engine he'd work on. Unless the company in question is the BigWorld, which knows a lot about the part of the engine. But Bigworld needs to get some skill in rendering. So it's the same either way."

SerB also mentioned the new graphic mechanism that already is in WoWp will be implemented into WoT - a scraped off paint on tanks, the metal under it "shining through".