@c1 hardware can't be fixed in software
Yes it can. Since I'd say, the end of the 1980's the division between software and hardware has become blurred. A certain set of functionality can be implemented in hardware or software, for example, a digital filter can be implemented as a shift register made from flipflops with feedback paths. The exact same function can be implemented as a software module in a high level (or even low level) programming language (and often is).
It follows that a problem in hardware can (but not always) be fixed in software.
In fact from around the end of the 1980s/early 1990's when Xilinx introduce the RAM based field programmable gate array (FPGA), the hardware design is actually a software issue, with new 'software' being downloaded into the FPGA chip to change the hardware design.
I'm not denying that sometimes the proper hardware fix to a hardware problem is the 'best' solution. I encapsulate the word best in quotation marks because the question here is what is the definition of the word best? What are the criteria?
The hardware fix might be the best in terms of performance, but in Apple's case, as with any business, money is a very important factor.
If enough performance can be gained by fixing issue in software, a hardware modifiction might be the proper solution to a hardware problem, but that requires a redesign, product recall and product re-issue, the 'best' solution as far as Apple and money goes is to release a software mod.
I'm pasting here a post I made in relation to a similar news item so you can see that a software modification could very well be a viable solution to this hardware problem.
"Actually, they could fix a hardware problem with software.
The article talked about signal processing and signal to noise ratios.
Fundamentally, the issue is one of the signal strength being reduced, and yes, you can't change that by software.
But, what they could do is change their algorithms for demodulation and decoding the signal to give a better signal noise ratio and enable the signal to be recovered whereas hithertoo the SNR was too low because the signal was too weak.
Putting in some noise reduction algorithms, or implementing some digital filters may help.
I'm interested to see how effective this might be.
It could turn out to be nothing, it depends on what the noise level and signal levels really are.
It could be that the signal level is way above the noise level, and holding the aerial simply reduces the signal level by a few dB and the signal is still above the noise.
In which case, adding noise reduction probably isn't going to help.
If when holding the antenna the signal drops below the noise floor then noise reduction and techniques to extract the signal from the noise will help."