i need community opinion about different approach i find. i call it double inverse model
the key is: no new routines and calls!

0. hw: plug MAP sensor instead of MAF
1. pterw:
  -pterw_w / pterw is the some kind of tank pressures with DSTGRAD / DSTOFS routine perfectly fit for DSS path
  -replace udst_w (pressure sensor) with uhfm_w
  -replace signed to unsigned overflow checkings
  -grad/ofs values should be set to make pter_w/pter same factor as other me7 pressures.

2. mlhfm: linear as the sensor curve are linear too
3. kfkhfm:
  - axe replace rl -> pterw with respectively axe values from 250 hpa to 2550 hpa
  - table values goes from 0.12 at idle to 1.8 at the high rpm to multiply MAP reading and simulate MAF sensor
  - the values at the pterw axe are flat for now, to fix VE model probably
 
4. log pter_w vs ps_w and mlhfm_w vs msdk_w and fix kfkhfm smoothly  <- im here
   - the models looks promising, i see ps_w pretty close to sensor readings pter_w.

so the pros:
 - pretty easy convertion, no need for new variables, rotuines etc
 - LP filter already implemented
 - hfm diagnose absolutely applicable here
 - failback to alpha-n (hi Nye!) is here in case of sensor failure

cons:
 - pter_w routine called from 1000ms raster. for now its irrelevant due to kfkhfm values are the same for all pterw readings but in case i want to fix that additional call should be placed somewhere

questions:
 -should i cut mste (tev) from frloh_w calculation?
 -sould i touch KFURL / KFPRG?
 -what about DV, is it some kind of interventions i should worry about?
 -what about all of this?

 logs attached (its 5120 halved pressures)