Yeah, now the six year old is confused.
I'll take a crack at this, in a manner where the 6yo will cry for Mommy.
If goatpig or anybody else would like to correct me, please speak up. (I've never used BC-XT and am shooting from the hip in some regards.)
For now at least, the BC Core vs. BC-XT scare is highly overblown. XT is essentially a copy of Core with a few minor changes here and there. Armory uses Core only to communicate with the Bitcoin network, and only to read the blockchain data that Core stores on the computer. That's it.
Unfortunately, I believe XT and Core both store their blockchains in the same location. (I could swear luke-jr or some other contributor pointed this out on Reddit or the dev list yesterday. I can't find the comment in question, though, and could easily be wrong.) If that's true, there could be a problem once the chains fork if people keep switching between programs. I'd argue it's important for XT to store its data elsewhere if they end up forking, otherwise it'll be a mess if people go back and forth.
When will a fork happen? It essentially boils down to when XT starts allowing blocks that aren't acceptable under Core. Right now, that doesn't happen. XT does a couple of behind-the-scenes things differently but the end result, the blockchain, is exactly the same. If the increased block size is committed to XT, I don't think that'll kick in right away. Like the soft forks that have rolled out, X number of miners will have to indicate that they can handle the latest transaction or block version. When P2SH was added, the standard was ~75% for the new rules to apply to those using them and, more importantly, ~95% for the new rules to apply to everybody. (Will this apply to the megablocks? I assume so. Gavin would really be playing with fire if he just shoved megablocks down everybody's throats overnight.)
In other words, until the megablocks hit ~75% on the network, they
must be validated using the pre-megablock rules. Unless Gavin & Mike have some grand plan to flip tons of miners overnight, this will take months to occur. Even relatively non-controversial changes, like the soft forks,
take months to occur. There's no way this change will take place quickly.
What does all this mean for Armory? The blocks will be validated but it'll be quite awhile before Armory
must be upgraded. Even when it is, we can always add some sort of switch asking whether or not the user is on Core or XT. Once Armory knows what to expect, all will be well.
Does that answer your question?
Dollars to donuts I'm wrong on one detail or another but the gist should be accurate. I can definitely say that the devs aren't sweating this. We have far bigger fish to fry before we worry about the relatively trivial changes that would be necessary if the Core/XT split becomes a real problem. I'm reminded of OpenSSL's Heartbleed bug and how everybody was freaked out about it, even though Armory doesn't use OpenSSL. It's an issue for some people, yes, but it's not an all-hands-on-deck kinda deal for us.
EDIT: Also, as best I can tell, XT had its last commit at the end of 2014. There have been months of commits to Core since then, including the BIP 66 soft fork that's rolling out right now. Mike & Gavin have a fair amount of work to do in order to catch up with Core before they even think about rolling out the hard fork.
