1) Transformer Loss -
Any LED Driver will also have loss...any electronics will have loss, period.
My point is that a fluorescent drop-in is gonig to have double losses -- losses from the ballast used to drive the fluorescent, and in the bulb itself, if the way I understand the product dropping it into a fluorescent ballast) is correct. A well-designed dedicated LED ballast (essentially, a rectifier with some kind of current sense to prevent overdrive) would have very minimal losses -- less than a fluorescent ballast of any kind.
2) Heatsinking - it is in an alluminumn extrusion and very well heatsinked.
Many fluorescent fixtures however might trap in heat, which could be a problem, though you are right, an extruded aluminum tube would be better than most LED products sold now, many of which are enclosed in glass.
3) Focusability - There is no loss of focusability
It doesn't lose focusability compared to fluorescent, but it doesn't gain anything either. Say you want a downlight -- rather than using these with reflector, LEDs with optics pointed directly onto your workspace would be more compact and distribute the light more efficiently.
4) Lumens - I have four of these bulbs in my office and have used a light meter on them....they do in fact put out as much light as a fluorescent that is rated at 3000 lumens.
5) With the ballasts that are in my over head fixture, it is running at 25watts, not 32...so it is more efficient than fluorescents.
3000 lumens from 25 watts? 120 lumens per watt is significantly higher than the specs on ANY LEDs on the market right now (and that's based on data sheets -- in the lab, with no fixture losses, with ideal heat sinking) This may be an exceptionally good product, but I'm skeptical that it's
that good.
If this product is even half as efficient as that, I'd love to see this company start making some dedicated fixture. Almost all other LED products after driver losses, heat inefficiency etc. are around 50 lumens per watt, or less. Worse than a CFL at 60 lm/W.
6) They cost me $149 per bulb, not $2 (?) and they are rated for 50,000 hours.
$2 is how much a decent 5000K T8 lamp costs me. Two 32w tubes in a high-efficiency reflector in my garage is putting out as much light as the four old 40W T-12s they replaced in a typical (poor efficiency) fixture.
That's why I think fluorescent will be the last thing replaced on a wide scale with LEDs. While the LEDs are much better than a lot of older fluorescent systems (especially ones that were flickery, noisy, low efficiency etc) the best fluorescents are very high quality by comparison, have none of those drawbacks, and still very affordable.
Where LEDs have the potential to win is by taking advantage of their inherent features in dedicated fixtures. One Idea that I've read about is to use essentially blue LEDs, and then a translucent panel containing red and green phosphors to make a luminescient ceiling-panel fixture for offices to replace bulkier fluorescent panels. Separating the emitter and the phosphor allows for a lot of design creativity not possible with fluorescent (as a UV-emitter with no built-in phopshor would be dangerous).
Along the same lines, common fluorescent back-lit signs could instead be side-lit with blue LEDs, and emit different colors using different colored fluorescent paints -- much more efficient than producing white light in an inefficient fixture and filtering most of it, and it would allow signs to be much smaller.
