Originally posted by kadajawi I think it has to be done within reason, and taking into account the abilities of the sensors. Lets say you have a sensor that can take light rays at extreme angles, just like film used to be able to (digital sensors usually require the light to come straight onto the sensor). You don't need to have lens elements that straighten everything. On an old sensor, this would give you massive vignetting, but not on a new sensor. Lets say you have a sensor that can have a really high number of pixels without losing low light capabilities, as there is no space between the pixels. Such a say 40 MP APS-C sensor would give you the same result as a 16 MP APS-C sensor, when you scale down the 40 MP to 16 MP. Vignetting should be within reason, of course, so that when shooting high ISO the corners don't look much worse than the center. Luckily the sensor is pretty sensitive in the first place, so that those situations rarely happen. But if you can have significant improvements in other areas, would you allow this vignetting? Since the sensor has 40 MP, the worst distortion can be fixed, and you'll still get a really sharp 16 MP photo for example (and do you really need more?). We are moving in the direction of such sensor tech.
I think one big question is backwards compatibility. Are you willing to develop and produce lenses that will look bad on old cameras, but good (enough) on new ones? IMHO the result counts, and price, size and weight do play a role. If you can leave away a couple of lens elements and in turn get a nice, small lens that performs pretty well, and the end result is good enough... why not? Such a lens would be a nice walkaround or kit lens. There's a market for high end lenses that try to deliver the absolute best there is, but even they can perhaps profit from corrections. What if, if you leave away something that fixes what could be fixed in post without any drawbacks, you improve the performance, sharpness, etc. in other areas?
I don't design lenses for a living, I'm just speculating here. But I don't think lenses designed for electronic correction have to be a bad thing. They can be, but they can also give us advantages.
Honestly theses are many very different designs out there.
The DA35 plastic wonder that inherit from older FA35 (that itself my inherit from older lense, I didn't check). This lens is both extremely cheap and achieving great result with reliable, fast AF, very sharp center sharpness, decent border sharpness, high resistence to flare, low coma/astigmatism. This lens is small, even smaller than many mirorless design and the FA version can cover FF.
The DA50/FA50 are a bit soft wide open but go as sharp as you can imagine closed down. Contrast and rendering are great. Still thoses lenses are small and the FA version cover FF. They are also quite innexpensive too.
The FA31, FA43, FA77 are all very small prime compared to what they provide and still are exquisite lenses with no many alternatives at the same level, all mount considered.
DA40, DA70 are pankakes, difficult to go smaller than that still they are really great.
In the end in Pentax world we already have small lenses, typically smaller than what many mirrorless provide, in particular FF mirorless. It is only for APSC wide angle that we have to compromize on border sharpness (DA21, DA15) but the lens are really small already and offer great rendering and great resistance to flare to compensate.
Why should we accept worse overall design than theses lens, get lower quality overall, need more post processing... get price just as high or even more expensive (look at Sony FE lenses prices) and still have to cope with bigger lenses in the end ?
That's a bad trade-off because you simply get nothing in exchange. Bigger lenses, bigger price, lower quality.