You're missing that the profile view on the FAA charts is not to scale, and regardless of depiction, the altitudes depicted are only on glideslope if they have a lightning bolt or are a glideslope check altitude like the small "2500" above the FAF. 3600 is not glideslope intercept at HUBJU, nor is 3000 at FIKOX. The line preceding the FAF is notional, representative only, and is not to scale or used to indicate anything about the glideslope. It's not the greatest depiction, especially since the glideslope feather extends well up final.
This is clear when you think more about it, though - HUBJU to FIKOX is 600 feet over 4.2 nm, and FIKOX to BLKEY is 500 feet over 2.9 nm. Obviously those are different slopes, and ILS glideslopes don't bend like that. A 3.0 degree glideslope like published here is about 318 feet per nm. So the glideslope altitude at FIKOX is about 2500 + (2.9 x 318) = 3422, and the glideslope at HUBJU is about 3422 + (4.2 x 318) = 4758. In other words, well above the charted altitudes, as is desirable to be able to intercept the glideslope from below.
Not flying IFR for two-and-a-half years has apparently made me rusty at reading approach charts! I was misinterpreting the stepdown altitudes as glide slope altitudes.
So if he was at 6000 MSL two miles outside HUBJU, that gave him about nine miles to reach the 2500 MSL glide slope intercept altitude by BLKEY. That's about 389 feet per mile, which shouldn't require a heroic effort.