If the gate valve was an open pipe then it would better correspond to Keston's secondary pump scheme, so I don't see that as a problem.
The gate valve will have a very small hydraulic resistance; the 'gate' should be out of the water stream when the valve is open. The LLH still works as Keston's recommendation, albeit that the pipe should be larger.
Apart from extra pumping, no energy is wasted; half the flow twice the delta-T = same energy.You are wasting 50% of the boiler energy to keep the delta T from going off the scale.
It lowers the efficency, I'd think by about 5%, but you can't avoid that for short periods if you should need 75degC flow for a HW demand or to supply the space heating demand.
I suspect that most balanced modern domestic CH will use a by-pass to increase boiler flow once the TRVs kick in and increase delta-T above the safe limit for the boiler.
If you look at your temperature curve, the problem is that Fpri > Fsec. The excess primary flow goes back into the boiler return. If you increased the secondary flow, it would alleviate the problem. A weather compensation system will do this. 100% flow at half the secondary delta-T, but still at maximum efficiency because you keep Tr below 55 degC by lowering the flow temperaure. The boilers' delta-T is still controlled at 15 degC by the boilers' controllers. The secondary flow temperature is reduced by mixing in the LLH, since Fpri < Fsec.
The system will deliver the design flow at 70 degC flow at 100% load, i.e, on a cold day.
On a mild day, 50% load, the existing system will deliver 70 degC; the TRVs will throttle the flow down to 50% or less.
A WC system will reduce the flow temperature to the minimum required to supply 50% load and the flow rate will increase to supply the same 50% heat load. The flow temperature should vary to maintain the room temperature in the one room with the sensor. The TRVs in other room will only trim the flow slightly. You should still get 90%+ or so of the design flow rate.
OK, many systems are not balanced, so the total flow through the radiators can be more than the design flow for the boiler and the by-pass need not open. Even when the TRVs reach their set points, the delta-T from the radiators might still be lower than the limit for the boiler. But you wouldn't leave radiators not balanced, would you?
The LLH system disconnects the boilers and radiators. The flow through the boiler is set by the pump speed and the boiler's controls. The flow through the rads can be one tenth of this, or ten times this and the boiler flow rate is unaffected. The boilers' delta-T is controlled at 15 degC by modulating the burners.
An infinitely variable pump would be ideal but with a suitable analogue controller would be much more expensive. The 3-speed pumps provide adequate control and are cheap, being a standard domestic CH pump. It's a clever system.