The I2K was invented in an attempt to increase the lifespan of the early PTFE couplers and to increase the maximum operating temperature of the machines.
With the TFM, the lifetime of the spacer increased dramatically compared to the earlier PTFE and glass filled PTFE.
For normal printing (up to 260C) I see no need for I2K if you are using a TFM-coupler.
If you for some reason want to go above 260C, I2K (or a PBI-spacer) is motivated.
Recent years has seen new plastics that prints at lower temperature though, such as the Ultimaker ABS that now prints at 230 compared to 260 at the time the UM2 was launched. So there is much less need for high temperature solutions now days.
Regarding the rigid spacer, the original spring-design has a flaw that causes trouble in the event if a leak between the coupler and the block.
If this happens while you are extruding with high force, the hydraulic pressure in the molten plastic can overcome the force from the spring, lifting the whole spacer the same way a hydraulic cylinder works.
Here is a very old picture, from the time before the Olsson block
, of what happens:
The adjustable spacer was a little design challenge to enable dual printing experiments on the UM2.
I recently made a fixed printable spacer with a slot were you can see the bowden tube, I think it is the best compromise for single extruder printers, haven't had time to refine the design or to upload it yet though.