An E-190 Captain (and Check Airman) believes the company's SOP for structuring and navigating the published FPA (Flight Path Angle) on the BOS LOC Runway 4R results in a descent below both the ILS GS and the PAPI. He suspects the published FPA may be inaccurate.

Narrative

Flying the LOC 04R Approach according to the published approach procedure and our company SOP we were well below the glideslope before reaching the DDA (Derived Decision Altitude).I believe the charted procedure is incorrect. We flew the ILS with glideslope inoperative a month ago and were below the correct glidepath. So I have since performed the approach in VMC four times and have found that the present procedure is more of a 'dive and drive' procedure that [the] FAA has been trying to get rid of. The glideslope is 3.0 for the ILS; but supposedly 3.2 for the LOC (which crosses MILTT about 80 FT higher than the glideslope crossing altitude). Obviously not scientific (note the inconsistencies; probably due to flying and writing); but here's my data for our four approaches using 3.2 FPA (Flight Path Angle) from 0.4 DME prior to MILTT: The charted altitude at MILTT is 1;800. On four approaches we crossed at: ?; ?; 1;760; 1;760. At six DME charted is 1;500. We crossed at: ?; 1;360; 1;460; 1;400. At five DME charted at 1;160 we crossed at: 1;060; 1;000; 1;100; 1;040. At four DME charted at 820 we crossed at: 700; 680; 760; 700. Maybe if I hadn't started down 0.4 prior to MILTT (per our SOP) it would have helped slightly; but the 3.2 degree glideslope definitely seemed to put us slowly further and further below the ILS glideslope. I had to take over manually at the 3 or 4 mile point each time due to being very red on the PAPI and hearing 'glideslope' too many times when the glideslope was working. If the FAA truly wants a constant path descent; especially from the DDA/MDA to the runway; then they should flight test or do the math and adjust the FPA for this approach.