If a little 6-pin part requires a quadrillion-gate FPGA to support the ISP task, well, that has to be figured into the cost of using that device

now, again, you say:
"If for MY (the 100 chip/year user) convenience, those that use millions of chips just have to pay a bit more, what do I care, after all MY convenience is all that matters".

It is far from Denver to RTP, otherwise I am sure I could hear you stamp your little feet in the floor.

To stay in business all chip designs (save the $1000+ superchips) must be made to the dogma: "anything that cost money, and hss no advantage for the million chip year user, will be excluded".

The million chip per year user do not give a friggin' hoot if the ISP is 'standard' or not. Thus ISP will be implemented in the way that is the cheapest to include on the chip.

I am not arguing with your preference, just your lacking sense of the reality that no chip manufacturer can stay in business by listening to the small user.

Erik