Spent yesterday afternoon, and this morning making voltage and
resistance measurments. I know know a whole lot more about 1990
Honda civic ECU.ECM than I ever wanted.
Only by comparing voltages from my 1991, and the "dead" 1990
was I able to eventually decipher the mess.
The return/data voltage from the MAP sensor was fluctuating
wildly. From about 2V up to 4.2V. and the variation only
happened when the engine was turning.
I found both the yellow/red(data +5.0V) and the green/white
(data ground/return) wires were damaged about 1' from the
MAP sensor. Iboth wires looked as though they had been pinched,
but the outer jakced was undamaged.I started at the MAP and
just flexed the wire and when the voltage  went "nuts" I knew
I found the spot. Stripped each damaged wire, made a "Western-Union"
splice, soldered and covered each splice with some conformal
heat shrink tubing.
I reconnected the battery, and the engine started with minimal
cranking. Much faster then it ever had.
We drove it areund a bit, and when I checked under the hood,
I found a small oil leak under the valve cover above the
distributor. I suspected we would have to replace the valve cover
gasket, but I was hoping to avoid it.
Thanks to everyone who offered advice.
We will change the sparkplugs when I am done with the gaskit job.
I never got up with my friend with the deep reach compression gauge,
so I improvised. I bought four "no oil foulers", those gizmos that
have male spark plug threads on one end, and female on the other
with a 3/16" hole. By stacking them end to end, and sealing with
locktite, Iused a lot more then I would for a normal applciation.
By screwing it in, I have a 5" extension that my short compression
gauge would fit. I compared my engine to hers, and while the extension
did lower the absolute values(figure a 1/2" bore, 5" long). All eight
(four each) of our cyls matched to within about 3 PSI.
This has been a bear! I would love to have the spcial test harness
that Honda specifies, it woul dhave helped a great deal.  
Terry