Yeah. I do not have any right to demand anything. However many spy camera advocates say that no problem, just say that you don't want to be filmed and we stop. No problem. And I refuted that this is not a viable course of action.

There are many things in life that happen because people do good things to each other without it being a right that can be demanded. I cannot demand much of anything throughout my day but still I like everybody else get hundreds of things my way every day even without asking. For some reason some spy camera people don't understand common courtesy and seem to only work based on what others can or cannot legally demand from you. That is one more reason why I don't trust spy camera people to do good things to other people.

In nowhere I said I am okay with security cameras. I do not like when there gets another bigger problem on top of a smaller problem.

I do not want constantly look at some strangers face whether there is some LED on his spy camera or not. I have other things I want to look at and concentrate to. Therefore the LED solves nothing about this privacy problem.

Let's imagine a situation if I sometimes notice the LED is on meaning there is a recording ongoing the only option I can do to avoid getting livestreamed to internet is to leave this restaurant where I am having a dinner. I cannot go and say to the camera guy that I don't want to get recorded because of Streisand effect. To avoid this problem the only sensible thing to do is to leave this kind of restaurant immediately before ordering anything to eat if I notice there is somebody with these glasses on. It makes zero sense to stay and constantly monitor whether the LED turns on at some moment and only after that leave the restaurant without getting eaten my meal that I had paid for.

All of this is something the camera guys do not think about. They just want to have a camera pointing to other people and say no problem, there is some LED that makes all of this okay. The LED solves nothing.

You might had chance if you said you target some esoteric high end professional equipment manufacturer. However you said you target a consumer equipment manufacturer and for me that says you have snowballs chance in hell with this plan. Even if you already had a physical chip in hand rolling out from a production line they are not going to add your chip to their board. It makes zero sense to have additional chips for some video processing algorithm. Their product already supports tens of similar algorithms and those run on a processor that already is in the board. If your doohickey can not run on the same processor it is not competitive. Let's however think about your question. I used to design ASICs and FPGAs but that was decades ago. What I am going to say based on that is even more true today than what it was in 1990's. First the ASIC designer must decide the general architecture. How much HW or SW is inside the new ASIC that we are making. Is everything written in VHDL or do we just put some CPU and bunch of memories to our brand spanking new ASIC and run everything on our CPU? Your question very much smells like the latter. Let's just splash a CPU and bunch of memories to the ASIC and run it in SW. Now the question is how fast you want to go do we put a small or bigger CPU? Your equipment manufacturer already has CPU and memories on their board. They do not want some new chip with CPU and memories on it on their board. If for some reason this is not true then you need to convert your design into VHDL or Verilog to be implemented in logic gates and not on that CPU that was my first, second and also third most logical choise for implementation architecture. Converting the C into VHDL can be done automatically but that automatic process results in so much unbelievably unpractical VHDL that it makes zero sense to make an actual ASIC with it. Converting C to VHDL in any way that produces a sensible ASIC requires a complete redesign of the algorithm. C is for serial processing. ASIC HW needs to run in parallel. The algorithm must be written from scratch to run on thousands of logic gates clusters in parallel. Many algorithms do not want to bend into this redesign at all. Whether your algorithm can be parallelized you need to hire somebody to do that to see if they can get it to work. Put up an ad. "Professional ASIC designer needed !! We pay well !!!" If they can make it happen the result will be so small in todays standards that the gate counts (which is a number used to describe a size of an ASIC) etc. do not matter at all. The gate count is irrelevant to the question. The ASIC is going to be way to small to be economical to produce. If I am wrong and the gate count is bigger it is even less economical to produce.

The amount of other phones in the cell affects what are the most important characteristics of phone to optimize for. The issue in here is about the minimum power levels which mainly affect the radius of the cell. However as the amount of other phones have gone up the situation where there is an empty cell where the radius could be an important factor has gone to be not so important any more. Nowadays it is all about signal to noise ratio at higher power levels that matters in a busy cell and that is not what was studied in here. The issue about reduced radius of a cell has been fixed in network design by putting up more basestations which are needed in any case to support the increased amount of phones in cell. So to summarize. Yes the phones are worse in this respect but this particular performance metric is not so important any more as it once was and phone manufacturers have allowed it let it go worse as nobody really cares about this one any more and other things are more important now.

The word telepathy has existed for hundreds of years so the word and its traditional definition have a point
whether it is useful form of communications to you or not

Telepathy requires some paranormal method of communication.
Something that is not possible within the laws of physics as we currently know them.
So building any apparatus however small and however close to brain, it is still not real telepathy.

Does putting a man-made cone on the nose of a horse make it a real unicorn?
Quite definately not.
And that does not depend on whether real unicorns exist or not.

Hope you are not offended if some stupid journalist decides that he is free to redefine the word "hacker"
if its original definition does not sound useful for him personally.