Thinking back to my first job as an EE in the early 90's, I clearly remember the new owner of the company who had been a CFO in his previous life, trying to figure out a way to move our hardware sales to a leasing-type model. His experience as a CFO led him to the conclusion that larger companies would prefer to bill items to OpEx rather than CapEx for just the reason you said.
It's more that the need manifested itself with the economic crash, as companies realized they would rather keep as much cash as possible in the bank in order to more successfully navigate the uncertain economic climate. Subscription model satisfies that risk aversion.
In some companies there are also budgetary reasons. For example, a department may blow their capex budget on something early in the year. If you have a payment model that supports regular payments, the customer can still buy your product, as opposed to having to make room in next year's budget (which is not only not guaranteed, but also delays the purchase).
But the biggest reason is ROI. Let's say there is a software that will save your department $1000 per month. You can either buy this software for a $24,000 one-time payment, or for $500 per month. If you do the former, it would take you two years to start realizing the ROI. If you do the latter, you start realizing it immediately. And of course, if you don't like the software, you can always cancel your subscription, which means you're also protected against the risk of, say, buying something that never gets adopted by the users (or project failures, etc.).
As it happens, I would have been fine just dropping the cash. No major emergencies occurred. But I didn't know that at the time.
Opex is much more reactive, both in accounting and in scaling costs of goods with actual sales.
The reason is simple: one is a capital expenditure, whereas the other can be categorized as an operational expense. With a subscription model, even if you end up paying more in the long run, you realize the return on investment almost immediately.