But, generally this will be to a much more predictable level that if you are relying on upfront sales. Unless you really mess up your product, if someone subscribed last month, you can generally predict that there is a, say, 95% chance that they will still be subscribed next month.
If it is a one off payment, then you know you aren't getting any income from them and you just have to hope your sale people are going to do their job next month.
It has nothing to do with lock in. That's a bit of a red herring. You can have just the same (or more) chance of a repeat subscriber by creating a really good product as you will with some kind of hostage strategy. As you say, not a winning business model.
My experience here comes from more business to business where a months worth of expenses would be covered with 1-2 sales. I had times when the sales people would just disappear on extended unannounced holidays or decide to move to a different job without informing anyone. With our business strategy relying so heavily on upfront sales to maintain the cash flow, it was not a fun position to be in!
Things may well be different with consumer products where sales volumes are in the thousands of units.
You realise that your post sounds like the stereotype of every bad manager ever? "How much harder could it be to build the new feature if we did X, Y, Z? Surely you can get it done in just three days?"
And if you get a reputation as a hostage holder, you will likely spawn competition that avoids your negatives.
I’m sure you will get some residual low hanging fruit in the form of people that buy subscriptions and don’t care enough to cancel.
Doesn’t seem like a winning business model though.