Some of the comments on JSF, Gripen and the Dutch made me realize that I'd been talking about the same stuff at the Aero-India seminar last week, which was indeed held in a conference center attached to the National Institute for Mental Health and Neuro Sciences (NIMHANS) in Bangalore.
I have loaded the file at http://www.sendspace.com/file/z2v3l1. (Ignore any pop-up ads).
To boil the theme down: I started by pointing out what fighters do that other weapon systems can't, and it's a pretty long list, from strategic precision strike with standoff missiles, to close air support and intelligence, surveillance and reconnaissance. Furthermore, they are highly mobile and fast (a 600-mile radius within an hour or so is a big piece of real estate).
But another point is that although they're expensive, fighters are affordable - because they last a long time. There are a lot of fighters out there, the Dutch F-16AMs being an excellent example, which are more than 25 years old and still doing a useful job. In case you haven't noticed, there are a lot of decommissioned warships younger than that.
Fighters last this long because they are adaptable - both to absorb new technology and adapt to new strategic realities. Nobody dreamed when the Dutch bought their F-16s that guided weapons would get so cheap, or that targeting pods would get so good, but that's how they do a CAS mission today.
I then made the point that stealthy-versus-not-stealthy is an oversimplification.
Taking off from that point, I observed that LO is not free:
So the questions remain: will reduced RCS plus jamming get you into most of your targets? Will that be more or less robust against developing threats than stealth? For the toughest targets, how many can you handle with cruise missiles?
How much diversity do you need in weapons loading? I use this USAF photo of an operational F-15E:
It's carrying five different types of weapon (SDBs, LGBs, JDAMs, AMRAAMs and 20 mm rounds).
Another issue: how adaptable will stealth aircraft be? The cost of putting an antenna or an aperture on an LO airframe is undoubtedly higher; and any transmitting or emitting device needs to be vetted for LO compatibility - for instance, the next step in targeting pods may be active laser illumination.
I don't think anyone has the final answers to all of it, but it does show (to risk getting boycotted by PETA) that there's more than one way to skin a cat.