IIRC, the pandas provs fae came OOW with around 8k GCPA (it's not that hard, especially for a fae in war with high base peons -- 4 mana/tick and tog is massive). They were below median, so explore costs would have been pretty small, leaving them with ~20M gc before whatever they earned + togged in 3 days. For toggers, with 7 days prep (4 days eow + 3 days fort) you could reliably get to -5 PPA without exploring w/ a fair amount of dragon gc or for conflict gc. Cheap non-toggers could generally get to around -3. With exploring, -2 PPA was doable, depending on costs. Losing 1 day prep from EOW should mean it's still viable.
They had much less military needs than your dwarves did, being oow.
To be honest, it was the obvious play and you should have expected it. They had no reason to not dump their pool, and it should be recognized that for sub-median provs with at least ok cash, exploring pre war results in more total military even if less mil/acre.
for what it's worth, I'm pretty sure the two big mistakes by pandas thus far:
1) not training ospecs instead of thieves on those rogues -- sorry, two rogues by themselves aren't likely to bring down your cows that are likely stacked with WTs. With 3 OSPA they could farm land from your upper core reliably in war and gain peon space for tog.
2) Bouncing instead of breaking with razes/in range trads on your core. Actually, not entirely sure what their gameplan was there. A 99% bounce (and I'm assuming it's high send bounces) is like 1/3rd the kills of a break.
Biggest mistakes by spartans:
1) Not going HAM with land hits. You've hit for like 500a. If they had razed instead of bouncing on your core, you'd have lost almost all your peon space and be in really bad shape.