Interesting. A mate of mine also came up with an online MMO CCG concept. Naturally he wanted me to be his programmer in exchange for equity and well, that's about as far as things got.
Done properly, you can combine the variable-reinforcement ratio addictiveness of gambling, CCG deck purchase or WoW raiding with the massive cashflow potential of downloadable goods. Like CCGs it could be a license to print money, but with even lower costs.
Computer-only CCGs also open up design space that can't be done for real-life card games. Picture "destroy a random permanent" -- that would be impossible to do if you're sitting with cards in front of you, but trivial if it's on a computer. There are other game mechanics that are possible to do in real-life, but are annoying enough that they're not worth it. Imagine a mechanic that requires you to reveal your hand and take an action based on the average cost of the cards in it -- annoying to do if you've got a five or six cards in hand, but trivial on the computer.
In some ways it's a shame Magic: The Gathering Online has the same cards they print: I would love to see what the developers could come up with for computer-based mechanics.
The MtGO game was constrained by remaining backwards compatible with the table-top game, overlaid with that damn fool idea about exchanging the virtual and real cards[1]. That it worked at all is basically a software engineering miracle.
For me the real challenge, when going from tabletop games to computer games, is that in a table game it is trivial to have a rule-modifying rule; in a computer this is a source of substantial complexity.
Think about it: most of the player genius in CCGs is deck construction. That is, studying the rules, studying the cards, and mentally simulating the interaction of multiple rule changes.
Most computer games don't have rule-changing rules, or very few of them due to the cost of special-case coding. They generally resemble CRUD databases with some element of chance applied.
To me the central, and interesting, challenge for a computerised CCG is to capture that rule-modifying rule capability. This might be an area where the dynamism of Lisp is perfectly suited.
[1] It reminds me in retrospect of that camera Kodak produced which took film pictures, but had a screen on the back to show you what you'd taken. That is, a self-defeating and half-hearted attempt to deal with a disruptive change to a fat, profitable industry.
With a highly profitable paper card business, Wizards has always been concerned with cannibalizing paper sales and pissing off the physical card distributors. MtGO players are largely also paper card players - in particular, the intensely competitive ones who are willing to pay $4 per virtual booster for an online venue to test out competition decks - so matching the paper rules was essential. At the same time, the way they avoided problems with the paper distributors was to cater to the hard-core players and ignore the casual market.
TL;DR: Wizards is too dependent on paper card sales to design a next-gen online TCG.
You are absolutely correct about their position, which reminds me very much of Kodak throughout the 90s. Kodak was tied to film. Even though their own staff foresaw digital photography, they were bound by their own economics and basically wished for digital to go away -- hence their heartbreakingly ridiculous Advantix Preview (the hybrid film/digital camera I mentioned above).
There were effectively cards that do 'destroy a random card'. [1] There are also several cards that result in searching the deck (and then reshuffling it) which require time/effort and yet were made.
Arguably one card (and then more cards in their joke sets) that doesn't actually randomly choose, which they won't reprint [1], and which they've banned because of the way it works [2].
And the cards that force reshuffling do require time and effort, but not in a way that breaks the game. They wouldn't make a card that required you to shuffle your deck five times a turn, but they could do so online. [3]
Done properly, you can combine the variable-reinforcement ratio addictiveness of gambling, CCG deck purchase or WoW raiding with the massive cashflow potential of downloadable goods. Like CCGs it could be a license to print money, but with even lower costs.