But it also works in the other direction. This is the fundamental problem also of PoW change propositions. At a certain point, you need to dump your original bitcoins. That takes a lot of balls, and you might as well make the wrong move. So thinking that the big whales are going to dump the original chain for any new chain, is a dangerous bet - especially if they want to rip off one another too in this game.
The smartest move is to keep your double coins, and wait until the dust settles. But if everybody does that, you get an even value split of the market cap.
The 148 chain can't reorg, so it's not a total wipeout risk scenario. Sure if 148 doesn't pick up traction, you can end up with a coin that crashes hard, but notice how even useless bitcoin unlimited tokens in bitfinex still have a price, so you could dump them for cheap and still get some out of it.
The problem is more probably that if it doesn't get traction, miners mining on it will abandon it, and with the slow bitcoin difficulty adjustment, this chain will simply stop, because too few blocks are generated. So the risk on the 148 chain isn't zero either.
(and yes, exchange IOU are independent of what block chains do, or even whether they exist).
Of course, if the separation is long enough, it would be prudent for the legacy miners to make a small HF change to make the split bilateral. Pushing out such thing would of course be accepted by all legacy coin users, who don't want to see their chain disappear. The simplest thing to do would be to push out a Core version with a recent check point in it after the split.