But you only got 50% on the test!

1. YOu dont get more power out than you put in!!!!
2. Remember the coil 'engaged' current does NOT pass through the diode

3. You know that the nominal (theoretical) collapse voltage is 10 x
input... so that is 150 volts (rounded)
- In FACT the rev voltage at 50 would work fine, because it will only see
15 VRev for all intents... but to be safe, always buy for instantaneous
application which would be 150, rounded up to 200 v

AND given the 10x voltage..
4. Doesnt it make sense that nominal collapse current is one-tenth the
input?
5. And when you look at the spec sheets, the nominal current rating is
for CONSTANT!!!
PEAK current .. for the 1 amp 1N40xx series is THIRTY AMPS
- NO current passes through the diode when clutch is engaged, and only
for a millisecond or several while voltage is removed.

So if you see a handy diode that's rated at 2 or 3 amps at 200v, and it's
got a brand on it... it should be more than enough.  Certainly dont need
a 6 amper.

    Ok, coil resistance is ~4 ohms, voltage is about ~12, so the coil
    needs to bleed off about ~3 amps.  The diagnostic literature says
    the coil should draw ~3.5 amps, figure at least 4 amps to be safe.
    So should a 6 amp 400 volt diode do the job?

Certainly, with a margin of safety.    As "Backyard Mechanic"
points out:  Diode current ratings are for CONSTANT current.
The 1 amp 1N40xx series can handle 30 amps of PEAK current.
The diode operates milliseconds each time the coil is shut off.
So if you find diode rated at 2 or 3 amps at 200v, it should
be more than enough.  Certainly don't need a 6 amper.

    Would that "snub" the collapsing field to protect the rest of the
    car's circuitry?   I am assuming that more is not necessarily better
    but at least not detrimental.

Certainly, with a margin of safety.

NOTE:   The voltage spike from the disconnected coil will try to
rise infinitely to maintain the coil current, until some current path
is found.   It could be through a suppression diode, a resistor,
a capacitor, an ionized spark gap between the relay points, or
a breakdown of insulation in the coil, wire harness or controller
circuit board.  Without protection the usual path is the relay
point gap.  Controller boards are built with protection against
voltage spikes, but sometimes one gets through.

Actually, because your existing diode is shorted and not open,
it is still providing voltage suppression by acting as a resistor.
The problem is that it draws unwanted extra current during coil
engagement, and it may eventually open up.

Good luck.

Rodan.