I think AN/APG-81 range performance is severely underestimated by most people and it most probably equals or outperforms even Irbis-E rangewise and is otherwise far superior radar system. This is my reasoning for this:
AWG-9 and AN/APG-67 range performance is well known. Both use slotted planar array antenna and TWT transmitters and have otherwise similar construction. AN/APG-73 also has very similar construction and it also sits directly between these two radars. AWG-9 had detection range of 213 km and tracking range of 167 km against 5 square meter RCS target. AN/APG-67 had a detection range of about 90 km and tracking range of about 70 km against same target. As AN/APG-73 (of late Classic Hornets and early Super Hornets) is directly between the two, it should have a detection range of about 150 km and tracking range of about 115-120 km against that same target. This performance is very likely as Swedish sources claim Ericsson PS-05/A radar for JAS Gripen can detect that kind of target 120 km away. PS-05/A has quite a bit smaller antenna than AN/APG-73 and thus will have shorter range as the two have very similar construction otherwise. Russians claim their Phazotron Zhuk-ME also has similar detection performance as PS-05/A and it also has very similar construction and smaller antenna to AN/APG-73.
Why did I mention AN/APG-73 and all those old radars? Because USN claims that AN/APG-79 for Super Hornet has two to three times the detection range of the earlier AN/APG-73. Same claims have been made by a number of airforces taking AESA radars into service, so it’s very likely true. This would mean AN/APG-79 can detect that 5 square meter target out somewhere between 300 and 450 km away and track it between 240 and 360 km away. Now AN/APG-81 has larger antenna with much higher T/R module count meaning it has superior output power and antenna gain. My estimate would give it about 1/3 longer range without any other improvements. That would give it detection range of about 400 to 600 km against that same target. Of course AN/APG-81 is about a decade newer implementing much newer technology, including new T/R modules. So it likely has even bigger advantage than that due to each T/R module having higher output power and lower losses.
Now another angle:
http://www.acq.osd.mil/dsb/reports/ADA391893.pdf
This is a report paper made 15 years ago by US Secretary of Defense Office Defense Science Board and they should really know their stuff about technology. They predicted then that a very conservative estimate for ground surveillance optimized AESA radar with average power output of 2.5 to 4.5 kW would give a detection range of about 230 km against 3 square meter target and a cued search (smaller search area) range of about 300 km. Now they used extremely conservative technology assumption of 15 percent efficiency for the modules and about 10W peak power in each module. Modern T/R modules have efficiencies at least twice that and 40-50 percent efficiency is pretty standard with much more powerful (15W to 25W) modules readily available. This along with higher module efficiency would mean that those earlier figures for APG-79 and APG-81 are very believable as even the APG-79 should easily have at least equal and likely significantly superior average power. APG-81 likely has a lot more average power available.
I believe Russian figures for Irbis-E and their other radars as they are achievable. Irbis-E is said to be able to detect 3 square meter target 350 to 400 km away within 100 square degree search area (which is very small search area). This means normal search mode detection range is likely about 230 to 270 km against that same 3 m^2 target depending on many variables. Those figures are because cued search usually has at least 50 percent longer range than normal search. Cued search is useful for targets which have been quite accurately detected by other sensors and is not that good for searching due to small search area.
Calculating all of the above together, we can estimate the following range performance against different RCS targets (numbers are rounded somewhat):
Irbis-E
0.0001 m^2: cued search 30 km, normal search: 20 km
0.001 m^2: cued 50 km, normal 35 km
0.01 m^2: cued 90 km, normal 60 km
0.1 m^2: cued 160 km, normal 115 km
1 m^2: cued 300 km, normal 200 km
10 m^2: cued 500 km, normal 350 km
AN/APG-79:
0.0001 m^2: cued search 32 km, normal search: 23 km
0.001 m^2: cued 55 km, normal 40 km
0.01 m^2: cued 100 km, normal 75 km
0.1 m^2: cued 180 km, normal 135 km
1 m^2: cued 330 km, normal 250 km
10 m^2: cued 600 km, normal 450 km
AN/APG-81 (conservative estimate):
0.0001 m^2: cued search 40 km, normal search: 30 km
0.001 m^2: cued 60 km, normal 40 km
0.01 m^2: cued 125 km, normal 85 km
0.1 m^2: cued 220 km, normal 150 km
1 m^2: cued 450 km, normal 300 km
10 m^2: cued 750 km, normal 500 km
This only shows that 4th gen fighters will be detected by all of these fighters at very long ranges with ease. Even most advanced 4th gen fighters would be detected quite a long distance away. It also shows that AN/APG-81 will detect Su-35S at far longer distances than Su-35S is capable of detecting F-35. This also shows that even Super Hornet very likely has quite significant detection/tracking advantages against Su-35. Even if we take the lowest end numbers for APG-81 (for example T/R modules which have been available for over 15 years for US radar manufacturers), it still likely has at least equal range performance to Irbis-E. These numbers also show that Irbis-E has the advantage against all current non-AESA radars.