The reverie daytona originally intended to re-use the stock inlet flow adapter at the MAF tube.
see here:
https://www.reverie.ltd.uk/Downloads/Elise_S2_exige_s2_111r_Daytona_Fitting_Guide.pdf
That's the installation guide that shows the white ~10mm radius with vane (and some dyno plots at the end). I used this flow adapter for a while, but removed it after it gave me trouble.
The trouble came because the inlet adapter is not firmly held in place, just friction; mine fell off and tumbled around in front of the MAF due to vibration. That was an exciting event - full-throttle acceleration, cross onto the second cam airflow increase, then the flow adapter starts tumbling in front of the sensor, then the engine bucks and stalls at clutch-in. oops.
As a gauge of load, I think the engine controls use either MAF or alpha-n on the low cam, depending on throttle rate, but on high cam, the MAF sensor is used absolutely:
Notes about 2005 Elise fueling and timing control
I think there is not a sufficient sanity check on the MAF reading, either, and the car even dynamically learns the alpha-n load table based on the MAF. So an error in MAF reading is a catastrophe. I unplugged the MAF sensor and nursed the car home on slow roads after the flow adapter fell off.
As I understand, the load measurement is part of what drives fueling, and as you expect the AFR is modulated in response to O2 sensor readings under closed loop fueling. Even if fueling is open-loop with respect to the oxygen sensor instantaneous reading, the airflow load is still being actively measured and fueling is being trimmed by what was learned during closed loop operation. The airflow learning and fuel trim learning can act to oppose each other, depending on the quality of the calibration. I wonder if that was really the best way to design an engine control system.
Regarding intake/filter flow restriction: since the reverie intake does not use the vacuum intake valve, i placed a pressure sensor on the intake plenum vacuum port, and compared it to the pressure sensor that is onboard the ECU measuring ambient pressure. At wide open throttle, the values were the same (within the tolerance of the sensor), so I thought that the intake and filter did not present a noticeable flow restriction. That being said, the sensor was not meant to measure small differences in pressure - it has a 0-3bar range over 10 bits (i.e. ~3 mbar = 1.2 in H2O resolution). In contrast, reverie published some airflow measurements:
Air Filter Data | Reverie lTD /
https://www.reverie.ltd.uk/Downloads/Reverie Ltd Bare Cotton Gauze Air Filter Flow Rate Data_Issue 3_24042020.pdf
I'm not too sure about the engine protection that the K&N filter provides.. fairly large particles can come through that mesh without a dust cover:
Cone Conical Air Filter - 152mm (6
Despite playing around with my intake and my other thoughts about this topic, I don't have any practical experience with optimizing airflow. Your plan for a straight intake with machined MAF mount housing sounds reasonable to me. I think that some of off-the-shelf air filters have a built-in radius/velocity stack too, in case you don't want to machine that. I would guess that the best behavior would be achieved with a nice, wall-attached laminar flow across the MAF sensor.
