Thanks to those who responded. Suggestions were to increase the
top length until Rradiation is 50 ohms or to leave it as is. In
both these cases it of course is necessary to add a series
capacitor to cancel inductive reactance. It was also suggested to
add more radial wires. That would (perhaps) lower radiation
resistance more and require a matching network.
I woke up at 2 A.M. thinking "AHA!" The answer is I need radiation
resistance to be low so that I get more current flow and a stronger
field strength. That means (in this case of an L antenna) I should
go for resonance which will keep the Rrad low and thus increase
my field strength. After all, that's what all the antenna guru's
try for by laying 120 radials around a vertical. It seems they value
the field strength gain over the (mathematical) antenna efficiency
gain of keeping the radiation resistance near 50 ohms.
But as one person mentioned the horizontal length helps for local
work so maybe shortening that more isn't such a good idea. Depends
on what you want.
Decided to use a series capacitor and leave it as is. It should
be a pretty close match SWR wise.
It also occurred to me this is probably just a rehash of the old
question of 'is it better to match an antenna directly or use a
matching network?' I guess that's a personal decision and I really
don't want to open that up again. It didn't occur to me yesterday
that is what this post is really about. Things often become clearer
the next day. Reminds me of a school lab where I stumbled on a neat
relationship between voltage, current and resistance. Spent an hour
proving it. Called the instructor over and he said "Congratulations.
You've just proven Ohms Law."