I don’t have the details to be able to say anything useful. But there might be a narrow window of tax circumstances in which 70% of the equity without AMT but with loan costs is better than 100% with AMT + marginal long-term taxes and no loan costs. (Such schemes make sense if you’re concerned about not being able to exercise your options on short notice after getting laid off. If you have the liquidity, however, set it aside, take the yield and hold form after talking to a CPA.)
Then you're at parity with never exercising.
Exercise, no successful IPO: lose $x exercise cost
Exercise, successful IPO: lose $x exercise cost, gain $y share sale benefit
No exercise, no successful IPO: gain/lose nothing
No exercise, successful IPO: gain/lose nothing
Service exercise, no successful IPO: lose nothing, maybe gain some AMT credits or capital loss carryovers
Service exercise, successful IPO: lose $x exercise cost plus interest, gain 0.7*$y share sale benefit
Your best best case is exercising yourself and getting the IPO, but you have to weigh that against the likelihood of it occurring and your personal risk tolerance for the $x cost to exercise.
I would never buy a lottery ticket, but I will always accept even .01% of a free lottery ticket.
I'm not 100% sure about this, but IIRC these programs are typically structured as a tax-free loan to you. If the shares end up worthless, the loan is then forgiven.
So while you may not be out the principle of the loan, or the AMT, the forgiven loan may be taxed as income at some point in the future.