Ok, so what if we were to figure out....

1. A pressurestat setting that would get your grouphead to the desired "idle" temperature off the thermosyphon (with, perhaps, an adjustable syphon restrictor so you could "tune" this)

2. An amount of HX tubing that is large/long enough out of the boiler (between boiler and inlet to group) that the water entering the group was COOLER than target.

3. A small additional heating element around which that tube could be wound, with PID controller, so that the water entering the group would be at the same temperature as desired, and which incidentally, the group is at.

I would think this combination would basically eliminate any sort of "hump", and wouldn't be TOO difficult to add, nor would it consume much power (since its just a "boost" from a "close but too cool" temperature)

What am I missing... and why hasn't this been done on any of the existing machines?

I think Jim Schulman came up with an idea similar to this some time ago. He wrote about adding a solenoid in the thermosiphon loop controlled by a PID that would open and close, effectively opening/closing the siphon loop to regulate the head temperature precisely.
I've wondered about adjustable thermosiphon restrictors like the one on the Faema Legend (only one that I know of) and why they are not more popular. They seem to be a great way to regulate idle temps without screwing with the pstat, so you could have all the steam you wanted and not require an 8oz flush. It would be a fun (but brutal) job to add one in to any machine though!

Genesis wrote:1. A pressurestat setting that would get your grouphead to the desired "idle" temperature off the thermosyphon (with, perhaps, an adjustable syphon restrictor so you could "tune" this)

2. An amount of HX tubing that is large/long enough out of the boiler (between boiler and inlet to group) that the water entering the group was COOLER than target.

3. A small additional heating element around which that tube could be wound, with PID controller, so that the water entering the group would be at the same temperature as desired, and which incidentally, the group is at.

I would think this combination would basically eliminate any sort of "hump", and wouldn't be TOO difficult to add, nor would it consume much power (since its just a "boost" from a "close but too cool" temperature)

What am I missing... and why hasn't this been done on any of the existing machines?

You'd be awfully close to AndyS's uber-modded Silvia, which has a rope-heated grouphead (vice thermosyphon), PID controlled boiler, and PID controlled preheater for incoming water to boiler (and who knows what else by now ). Of course that machine is a single boiler.

Some of these ideas are discussed in here, PID on a HX...this is my thinking, is it flawed?, see Dan's link in there too.

HB wrote:Have you read MOSFET's Six Million Dollar Expobar? He "de-HXed" it by adding a PID and insulated preheat coils.