B-zar: As you may have noted, i didnt say "will" because i dont know that any devices will be sold at all.
No, you just stated "sells", implying there is precedence or other evidence. Consider the difference between "Bright colors sell more cereal", "Bright colors may sell more cereal" and "Bright colors will sell more cereal". The first implies a known phenomenon or correlation that provides reasonable evidence for a belief that bright colors will sell more cereal. The second implies some existing rationale it could happen, but no evidence. The third implies some well known and reliably repeatable mechanism that forces more cereal being sold.
The comparison is something you equate to something that is perceptual and beside the point of the product. And then you exchange the product for something irrelevant to the topic.
International support exists on its own accord,
because its selling extra devices. If it didnt exist we also wouldnt have it, but id like to side with rationality and say its because international support means support of many things, rather than just one.
We dont know that any particular brand of cereal fares better or worse, since its an image-thing. There is no control-group cereal with just the colour changed on the market. When that stops being the case for international support and devices, you would have to look evidece not stemming from its rightful place in the market.
I think the burden of proof is contained within the notion.
If international-support sold less devices, which can be the case, its for the same reasons it sells more. It is incompatible between regions. So if you make more than you should of one thing, it can potentially sell less, because you have tied your available devices for sale, to an estimate that could not hold true. Since it has this inherent problem, its another reason why it wouldnt exist over time, if not for selling more devices. Correlation does not imply causation, but that is my assumption. Interestingly, you can tailor the default to be a non-incompatible as possible.
There is an important distinction between lumping international support onto English, being the default, and having international support within English.
One is less compatible with english because it adds foreign stuff. You can do that once, or all over the place. (option here being the fullest possible keyboard)
International English+option
regionalized-International English+option made into printed keymat
Dual-regional English+german
Regional English - option
Dual-regional+multi regional English+german + many foreign letters
Multi-regional English-option+ many foreign letters
From the point of view of having one default keymat, international is best, because then you can make regionalized-international, which is for international users equivalent to English+option for English-speakers.
Spanish is largely irrelevant to them, so you would have to do a less frequent language for Dual regionalized. It is not equivalent to international regionalized because Italian and German arent QWERTY. We can imply hidden AltGr level for this reason, since it is always preferable in some way for both camps.
Regional (English only) is not equivalent to international even for English people, because then the English dont get their "option".
From the point of view of having mutually exclusive regionally familiar symbol-placement between different keymats, i dont think its worth maintaining many software-layouts over, it hast the optential to cause problems. And also I assume there to be one default.
When the default can work with so minimal changes for international users, i think its worth giving up half the english symbol-placement for. Most notably because its not possible to preserve it without also losing numbers for all.
Edit: As support argument, see also
why dedicated letters always beat language-modifiers.
And
why dedicated buttons or dedicated diactrics, alongside full dead-key diactrics, is preferable to no dedicated buttons and dedicated diactrics by default.