Well, as a commercial endeavor, I would really have to wonder if there's enough of a market to sustain a business. I'm sure there's the odd OS/2 system still running here and there but nobody today is interested in expanding that footprint. At best you'll sell upgrades to people who are still using it -- a customer base that will only shrink over time.
If it were an open-source project, yeah "why not" is a fine answer.
I agree it's not a growth market, but there's a lot of OS/2 in places you wouldn't expect, like ATMs and point of sale systems, if you keep the price low enough you can avoid increasing the pressure to upgrade, and you can make a comfortable income for a while.
This explain why Santander's ATMs (4B ATM) have the same UI since 90's and can't send an email instead of printing on paper when you do some operation on it.
As a general thought, hooking up ATM machines to the general internet sounds like it would be a source of "fun" for years to come.
For your point though, most probably it wouldn't be the ATM that does it. The ATM would send/queue a notification to some other resource in the bank network, for that to do the whole email notification piece. But that should be possible today (or even years ago).
I agree, but usually when people post nothing more than "why?", they're not communicating that they're a customer or potential investor who is concerned about the long-term viability of the product or the company. Usually when people post nothing more than "why?", it's a lazy of saying "I think this is stupid", because if they did just come out and say "I think this is stupid", it would make their absence of a supporting argument all the more glaring.
If it were an open-source project, yeah "why not" is a fine answer.