I was talking about journalists who ignore exactly that.
I wasn't just talking about individual inflation. For example, when a company that makes international sales reports record earnings in dollar, while the dollar sunk compared to almost every other currency, is this just misleading. Compare the revenue increase of the biggest companies with exchange rates, they are probably not as healthy as they appear on first sight. Maybe they are not growing at all and just the dollar inflated again.
I talk about the value of the dollar, when the author takes the metric 2.00$ this is relevant. The official inflation rate is in almost all cases totally wrong, because it takes way too few factors into account. This is a separate topic, how the inflation should be calculated. I think just that currency exchange rates are a better way to predict inflation trends.
"The official inflation rate is in almost all cases totally wrong" is a nonsensical statement. "Inflation" is not a uniform concept, to quantify it you have to take a certain basket of goods (the definition of which depends on the purpose of the exercise) and their nominal values and calculate the difference (or rate of change) of samples at various points in time. Furthermore, "inflation" is only relevant for nominal values; the that was used in the OP uses real values, and so do most statistics, because real value is the only comparable thing.
Especially your latest sentence shows that you have a completely warped view of monetary economics. How can one "use currency exchange rates to predict inflation trends"? That's a completely nonsensical sentence, but it doesn't matter what you really meant, because PPP is already used to combat this exact problem.
I can't believe I'm defending him, but he does have - perhaps accidentally - two semi-valid points.
First: PPP and inflation are both calculated using a basket of goods - the exact basket differs between researchers or institutions. Typically you want your basket to be fairly broad, and reflective of the goods and services people actually buy. But the very poor buy a different mix of goods and services than the average person - even the average person in a very poor country. IF the price of the most basic staples has increased more than the price of other food, or if the price of food in general has increased more than the price of durable goods and appliances, then the PPP adjustment used by the World Bank for, e.g., Uganda MAY misstate the actual purchasing power of the very poor[1].
Second: If you think that official inflation rates are being tampered with[2] then you may look at other metrics than should drive inflation. One such is exchange rates. If the Canadian dollar is weakening against a trade-weighted basket of other currencies, then the purchasing power of the CAD is declining. (Or if you prefer, if a lot of CAD is being printed, then we have more supply of CAD, and lower demand, resulting in lower exchange rates.) In any case, this should be reflected in inflation. If it isn't then either there is a big untapped arbitrage opportunity[3], the inflation statistics are wrong, or (more than likely) there is some other factor at work[4]. So, no, you really can't use exchange rates to predict inflation, but you almost can. :)
[1]: In other words, it COULD be that the bank is saying that someone living on $2 a day is better off now than a year ago, because coffee makers are so much cheaper that it gives them more money to buy rice, which has only increased a little bit. But if the very poor only buy rice and not coffee makers, they could be worse off now since everything they actually purchase is more expensive, even if the "basket" is cheaper overall. The choice basket of goods matters, and it needs to be appropriate for the group being studied.
[2]: This can and does[5] happen. It could be as simple as fudging numbers, or it could be more complex - for example, my picking an unrepresentative basket of goods. In the US, the most common measure of inflation - CPI - excludes food and fuel costs. This is very appropriate for picking up monetary inflation, but less so for calculating the actual change in living costs people are experiencing. As you say, inflation is not a uniform concept; if a government pretends it is they can pick a measure favorable to them.
[3]: In this case, you could buy CAD using your local currency, purchase goods inside Canada for CAD, export them, and then sell them for your local currency. This would drive the CAD up, and/or drive the prices of goods inside Canada up, until the change in the exchange rate matched inflation. And people seeking such arbitrage opportunities is one reason why inflation and changes in exchange rates tend to match, very broadly, over the long term.
[4]: And there usually is; exchange rates are incredibly noisy, and influenced by everything.
Well yes one can argue about the basket used in the referenced study. I haven't looked into which one was used exactly, but in my experience of working with UN data (of which the WHO is a part), their measurement methodologies are usually quite sound and well-reasoned (of course you can always find something imperfect in the margin, especially for decisions made in the past and that haven't been changed to make it possible to measure change over time, but I'm talking overall). But that's quite a difference from the claimed and unsubstantiated 'totally wrong'.
For the second point, I don't think I'm fully understanding. Yes you can use exchange rates as a proxy or second-order measurement for inflation (measurement, not prediction, unless I'm missing something); but I fail to see the relevance of that for the current study, since we're (presumably, again I haven't studied the methodology and more specifically basket used to measure inflation/PPP in that much detail) talking about goods that are relatively little influenced by exchange rates (i.e., much of the consumption of the people on those 2$ budgets is local and based on subsistence farming and local trade).
I wasn't just talking about individual inflation. For example, when a company that makes international sales reports record earnings in dollar, while the dollar sunk compared to almost every other currency, is this just misleading. Compare the revenue increase of the biggest companies with exchange rates, they are probably not as healthy as they appear on first sight. Maybe they are not growing at all and just the dollar inflated again.