100 mhz has sufficient resolution for the bandwidth needed, 1 mhz does not.
The
change in speed of wave propogation changing from free space to a metallic conductor distorts
all waveforms but is most unkind to the shape of square wave data.
In a resonant length antenna like FM broadcast at 100 mhz and 1 meter, the penalty is at a minimum.
In MW broadcast at 1 Mhz, resonant antennas are seldom seen on either end. :
We can "make" the antenna tune as though it were a reasonant length, but this comes at a big cost to
the linearity needed to maintain the squareness of such data.
The general effect is that in such a compromised transmission/antenna system is to "smear" the hoped-for square wave into something smoother, at which point they won't decode.
In AM HD transmission, this requires high linearity and wide bandwidth ( not easy ) through tuning networks.
All networks create phase and other distortions.
In reception, AM HD pretty much only works in a brute-force manner, and in almost any meaningful, normal
distance expected for AM service, it quickly loses the "crispness" if you will, to decode.
In every AM loop antenna in a modern radio. there is no real RF amp/tuning/peaking.
This would be an actual HELP, because it make the bandwidth wider, but effective sensitivity is much lower.
And there simply is no escaping the fact that as any normal MW antenna is but a fractional size of the wavelength,
there is going to be significant distortion of the (hoped for) square wave in the (2nd) change in speed, from C, to 9/10th of C.
It's wishful engineering at best on 1 mhz.
You can sweep water up a hill if you are really, really energetic and tireless, but eventually it begins to seem unwise.