As you've pointed out in the past, I've come to use macros in ASEM51 to do what a compiler might do.
Most of us use a compiler for "what a compiler might do"

That was, of course, because compilers weren't available yet for micro's.
remember those days.
I wonder how long it took those that used compiler rel 1.00 to try it again. I remeber, for each and every malfunction, having to see "is it my code or the compiler'. Fortunately, for me, this experience was with a latecomers beta and I already had confidence in the Keil tools.

Now that they're common, there's really no reason one shouldn't use 'em when appropriate
WOW, that seems a new attitude of yours

That, of course, is because I've worked with systems that were started back before those compilers were available, yet are still in use today.
As late as less than a year ago I worked on such and, of course, stayed in assembler. The occasional post "we have this assembler product, how do I convert the code to C" is laughable. However, if such a product, to meet todays demands had to go through a 80% rewrite, I'd start from scratch in C. Actually, some time ago I did an "interface C to assembler" in order to add USB to a device. But, again, "new code" not "maintenance".

The fact that we are moving towards where the 100MHz 64/64k version only cost $0.01 more than the 50MHz 32/32k version makes the reasonable use of assembler less and less.

Erik