Cecil makes good points. He does indeed seem an intelligent being!
I would expand on some of those points
- Regulations
The US, EU and most other standards will
effectively ban all ordinary incandescent lamps
- including the touted halogen replacements, as from the US EISA 45 lumen per W end regulation standard (halogens typically 22-25 lumen per W)
Standards explained, and 10 American state repeal ban bill updates (legislated Texas) Light Bulb Clarity: New Electric Politics
- Savings
All lighting has advantages, energy saving is not the only reason to choose a bulb.
Besides, light bulbs don’t burn coal or release CO2 gas.
Power plants might.
If there is a problem - deal with the problem 
While individuals may save on some frequently used bulbs,
- noting that not all bulbs in a typical 45 light US household are used much, and remembering about breakages, lost bulbs etc -
overall US energy savings from a switchover are a fraction of 1%,
based on US Dept of Energy stats and surveys, referenced
Light Bulb Clarity: New Electric Politics
also describing more relevant electricity generation, grid and home consumption savings.
Even when individuals save more, society energy measures should of course be based on supposed society savings, unless interfering with individual lives is a priority.
And individual citizens don’t save as much as they might think…
So, re Dangab original post on Heat,
while 95% of incandescents is indeed released as heat, it is at least usefully radiated (when it’s dark it’s often cold!)
Light Bulb Clarity: New Electric Politics research referenced
Conveniently ignored:
80% of CFL and 70% of LED energy is more truthfully “wasted as heat”,
since their heat is internalized, giving a greater less predictable fire risk, especially with CFLs (http//ceolas.net/#li18eax )
- yes, incandescents can burn lampshades etc, but that is more noticeable and predictable.
Moreover, the “power factor” of CFLs alone means that they use twice the energy to what your meter says, as Sylvania/Osram admit in the linked factsheet
(many domestic LEDs also have power factor issues)
- but users will of course have to pay for that eventually.
Again as Cecil says,
Jevons paradox about using more of effectively cheaper usage, should be remembered.
Conversely,
energy saving does not necessarily save money either.
Electricity companies are allowed to raise prices, or be taxpayer-subsidised, for expected lower electricity sales as referenced on the website (California, Ohio. etc CFL switchover state programs, as in Europe, UK, Canada, BC etc)
Re **Dangab **on CFL manufacture energy use,
there is some data online for example from Osram purporting to show relatively low manufacture energy use.
However, that relates to assembly of manufactured components.
As linked, German Dr Stanjek in research for Greenpeace (so hardly biased) pointed out that the real energy use is several multiples of times more than pure assembly.
A real overall comparison would also take into account the energy/environmental cost of mercury mining, ballast component sourcing and manufacture, CFL manufacture in China using inefficient coal plant energy source, transport from China on dirtily bunker oil (C02, mercury releasing) powered ships, recycling energy cost and in some cases transport back to China for re-assembly etc, similarly as appropriate with LEDs.
Certainly, incandescent manufacture is not always local, but more easily so, also for low-tech startups.
Incidentally - since Cecil mentions it -
no power plant would be spared even on supposed savings, certainly not any major coal or similar plant- it relates to the constant base load from such stations.
It is a bit like saying “Hey, if we reduce pupil numbers, we can reduce the number of schools built!” - but of course 70-80 rather than 100 pupils in an area still means essentially the same school facilities are required (and here it is more like 99 rather than 100).