If you sell half as much stuff as others, it means you're costing the store customers, because presumably you approached close to the same number of customers as the other sales agents approached, but more of your customers decided to take their business elsewhere. This not an issue in a programming job.
Let's not "take this to the extreme", and instead be realistic. Your "ticking bomb" would be the first in history. Like I said, no drug is tested for 20 years before being released (and approved by FDA), so if there was such a risk, then with the sheer number of new drugs developed, you'd see these disasters left and right. Where are they?
Tobacco causes severe side effects which get progressively worse before the "big stuff" kicks in. That's a pretty good cue that you're poisoning yourself. Not so with Provigil.
If you want one that sneaks up on you, try fen-phen. No noticeable side effects and then you die of a heart attack.
As for modafinil, a quick look doesn't turn up any long-term safety studies (pub med is either down or even slower than usual today, which amounts to the same thing, so I might be missing something), except for one 9-week study. You could smoke for 9 weeks without being affected much.
It's not like it's a rare thing for drugs to turn out to be harmful. We'll just have to wait and see.
EDIT: given the lack of long-term safety studies, but also the lack of problems with short-term studies, I personally would use modafinil for finals week or a long car trip or something like that but not day-in day-out for years.
Let's not "take this to the extreme", and instead be realistic. Your "ticking bomb" would be the first in history. Like I said, no drug is tested for 20 years before being released (and approved by FDA), so if there was such a risk, then with the sheer number of new drugs developed, you'd see these disasters left and right. Where are they?