On the face of it, a CFL has lower TCO (Total Cost of Ownership) than an incandescent since it consumes less energy to produce similar lumens - even it it only lasts as long as the incandescent as opposed to its rated lifespan of 5000-8000 hours.
But reality isn't so simple. CFL's don't like overheating, short-cycling, environmental extremes, and can do funny things to the circuit they're connected to. I also gather that thanks to their non-1.0 power factor, their faceplate energy ratings need to be taken with a grain of salt - especially with the deployment smarter electrical meters that charge residential customers for power factor correction (I don't know what the power factor is of the average CFL nor how much it distorts their apparent wattage relative to their rated wattage). There are likely some other legitimate performance considerations I've missed out on.
CFL failure modes are a bit less predictable than incandescents - sometimes they simply "burn out," other times components melt, the tube goes uselessly dim, and occasionally they spark or burn.
There is concern about their constituent compounds and elements - although I find the paranoia about mercury and other compounds curious given the small amounts present and the prevalence of these compounds in other commonplace devices (which are not the subject of such paranoia).
The race to the bottom on price has hurt CFL's badly. When they were $10 (more more) each about a decade ago they easily lasted their rated lifespans and rarely experienced bad failures. Now that the target price is <$2 each, corners have been cut in terms of durability, performance, and optical quality.
While there's no accounting for personal preference, I feel that many consumers' expressed disdain for CFL's seems to be more a reaction to change than a true substantive objection. I will leave this subject alone in this discussion due to its highly subjective nature.
As fond as I am of LED lighting - even in the significantly compromised "LED bulb" footprint - I think they will take a long time to gain traction in the marketplace and the race to cut costs will hurt them in similar ways to CFL's. Some of the early "5mm showerhead" models that had placement in big-box retailers several years ago soured much of the market due to their overall shoddiness.
If we are to see lighting as a capital expense rather than an operating expense, then one needs to realize that the trade is upfront expense rather than variable operating expense. Assuming both options offer satisfactory performance, then the decision comes down to which option has the lowest total cost of ownership. If electricity were exceptionally cheap, then the incandescent would win due to its extremely low cost to acquire and negligible operating expense; with electricity not being so cheap, alternatives such as CFL's generally win due to their extremely low operating cost in spite of their greater initial cost of acquisition (in fact, a $20 LED bulb beats an incandescent for TCO assuming it lasts most of its 25,000 hour rated lifespan).
This is idea a tough sell even to sophisticated commercial operators who are often under pressure - both external and internal - to minimize capital expenses in organizations that have internalized eternal operating expenses, regardless of the TCO advantage. It will be a far harder pitch to residential consumers for "low thought" purchases such as home lighting ... whom will generally only pay more for better fixture aesthetics than performance.
A challenge with presenting lighting as an investment is that residential consumers really really want a light bulb to screw into their existing Edison sockets. This presents a challenge to nearly all of the alternative technologies. Linear florescent requires dedicated fixtures; LED and CFL do poorly in most Edison socket fixtures; the other arc-light technologies out there might screw into Edison sockets, but still require new fixtures and ballasts due to their specialized driving requirements, and even if they are self-ballasted they suffer from most of CFL's limitations. If people are willing to replace Edison-socket fixtures with purpose-built fixtures, then the notion of lighting as being an investment might catch on ... but I'm not optimistic.