Adam,
It is so frustrating trying to program and test a microprocessor when all that is visible are four I/O ports. MPLAB X can't even show the contents of the registers, or at least I haven't found out how to do this. I'm just about 100% certain it shows the program memory contents after doing an upload, because that is how I find the factory-set oscillator calibration bits located in the last word of program memory. I don't trust MPLAB X not to erase these bits, so the first thing I do is read them and write down the value for later reference in an EQUate directive for each PIC I use (three of them so far). It appears that there are much more sophisticated PICs with lots more program memory available, enough to allow MPLAB X to download a de-bug "executive" into high memory. And for these more complicated PICs there are de-bug modules available that can transparently execute the target PIC code in real time, with breakponts and such. And you can even program these larger PICs with high-level languages such as C, which I know almost nothing about. As you said, at work there were always software weenies available to roll code for us hardware guys. I learned just enough C to "keep them honest" so to speak.
Getting the sleep function to resume from where it left off requires some means to save in a user variable the address where you want program execution to resume. This could be a table indexed by another variable whose value you assign just prior to sleep. Then, having determined the reset was a result of awakening from sleep, you use the index to retrieve and load the address where you want to continue, either as an Interrupt Service Routine or a branch back to the code following the sleep instruction. That's a lot of overhead for the PIC10F200 (or PIC10F206) that is unnecessary with the PIC10F320 (or PIC10LF320). Just for grins, I may eventually try to program that for the PIC10F206 to see if I understand indirect addressing and pointer objects. Hmmm. I had plenty of trouble with that using C, so no telling what it will be like with assembly. All the software geeks on this forum are probably ROFLTAO right about now. I make no apologies for being a hacker: I will beat on the bits until it works, no matter how ugly the final code looks. I do draw the line at writing self-modifying code, leaving that to the criminal hackers now. I may have written a few scraps of self-modifying code in the past, but the devil probably made me do it. It sure didn't get documented as such.
I would surely like to have a CNC milling machine. It is a great toy! A CNC lathe is also a fine toy. We had both of these (for awhile) where I worked, complete with all the tooling needed for setup and getting things done. Well, almost all... you can never have enough tooling. These were not top-of-the-line models by any means, but they were adequate. My company sent those of us who were interested to Sinclair Community College to get trained on how to use them, but I chose not to go. A year later they moved both machines to another building for production machining, which made them virtually inaccessible except by prior arrangement. So much for prototyping.
I am not so certain that a PCB milling machine is the way to go for prototyping. I like the idea of replacing the cutter head with a small diode laser to either ablate a resist coating or, with a UV laser diode, expose photo-resist coating. No messy chips to clear away, no expensive carbide tools to break. I would also consider building a small Gerber photo-plotter with an aperture wheel for several different-sized pads and traces, but I seriously doubt that is going to happen no matter how large my collection of stepper motors. My wife would insist I build her a 3D printer as soon as she saw me gathering x-y slides and whatnot. <sigh>
My engineer son had some success as an undergraduate student at The Ohio State University physics lab using a laser printer to make circuit boards. I think this involved printing a mirror-image of the positive artwork and then transferring the toner image on the paper to a bare copper-clad board by means of a heated flat iron. After the toner re-hardened on the copper, he somehow dissolved the surrounding paper and then etched the toner-resisted board in ferric chloride. Some touch-up was usually necessary prior to etching, but I was semi-impressed with the results. Cheap, effective, and fairly quick turnaround. After etching, the toner came off fairly easy with a 3M Scotch-Brite abrasive cleaning pad.
Hop
