Typically, when the price is highly visible to the consumer, certain price points work best: 0.99, 1.49, etc. You don't price an app or a hamburger at 1.82. To achieve that, you "bundle" your price increases.
A similar aesthetic could be achieved by showing the attractive USD price, and the converted price in a smaller font, or even on a different page/view/whatever (although that's kind of shitty). I've seen this approach in a few places.
Not having done forex, my guess is that *most* of the time, the major currencies would drift within a few percentage points of each other week to week, so sticking with 0.99 would work better than 1.02 one day, 1.01 the next, then 0.98. 1 is psychological cut-off, going above that reduces sales.
If it was that simple, I'd be a Forex millionaire by now, as would many others, since buying each dip would be a viable strategy. Unfortunately (or fortunately, since I didn't go broke), I ultimately threw out about 6 months of development into a complex algorithmic trading system that targeted Forex. Bear in mind I did that as somebody that's not afraid to take risks; unfortunately, currency pairs swing wildly, even when no geopolitical events can be easily attributed to the fluctuations.
To get an idea, take a look at the 52-week pictures for a few currency pairs:
These are major currency pairs, and as you can see, they move wildly on a daily basis, let alone weekly or monthly. Sometimes there are obvious events such as Brexit, but most of the time it's due to innumerable factors and a dash of irrationality. Fixed currency conversion rates do not have a place in modern society, and are almost always a scam. What Apple is doing is wrong, and I'm sure at least some of the people involved know.
Every nonzero finite dimensional inner product space has an orthonormal basis. It makes sense, when you don't think about it.