Yea, there was still a "LH2 is the miracle fuel" mentality back then. The real sad thing is there was no domestic RP-1 engine to even go to to offer. The next gen RP-1 engines partially developed for SLI wouldn't come along until the late 90's. But imagine if there's been a TR-107 or RS-84 engine to offer for an EELV? Topped with Centaur or DCSS or something else, would have made a fine all-domestic Atlas V with a bit more performance. And if they had been selected winner of the EELV competition, then it would have been in production by the time ESAS came along, and would have made a very viable engine for a HLV (assuming Shuttle Derived politics could have been over come). As it was there was only Russian built RP-1 engines flying and any RP-1 HLV would have needed a new engine development. That actually wouldn't have been that hard because TR-107 and RS-84 had been partially developed and those programs only recently cancelled with the cancellation of SLI. But it did make it easier to put the thumb on the scale for Shuttle Derived.
Well they had options other then the RD-180 they could have brought the F-1 back in production using modern tooling or used four x RS-27s.
The former would have out performed the RD-180 by a large margin the latter would have been lower performance but Rocketdyne might have been able to sell the RS-27 for a lower price if production was higher.
The RD-180 was cheap simply because of the dollar to ruble exchange rates at the time and the Russians were hungry for any business they could get.
But the TR-107 and RS-84 were the engines they really needed for a next generation launch vehicle at the time.
The TR-107 in particular looks like it could have been a very affordable for an engine in it's thrust class and it's high enough thrust you'd only need one for a MLV.
In retrospect cancelling SLI was big mistake.
The RS-68 was the result of some interesting trades. I think it can be somewhat difficult to separate hindsight out of an look at decisions made in the past. While the TR-107 and RS-84 were good looking engines to us now they were in some ways the antithesis of what the EELVs were about. The collective wisdom at the time was that if you wanted to make a cheap launcher you needed a cheap engine. The way to do that was though to be reducing the number of parts, using an ablative rather then regeneration cooled engine bell, simple gas generator cycle, and sacrificing ISP to make a cheaper engine. Performance, reusability, and pushing the technological edge with a new combustion cycle were traits which would have pushed the development and recurring costs up. Using LH2 had as I understand it two main advantages, first is that the expertise in engine design was in LH2, and secondly that if the engine is going to have a lower performance that LH2 gives more room for it to be degraded while still having a decently preforming system. With a Vac ISP of 414 the RS-68A has a very poor ISP compared to other large LH2 engines but that is still much higher than any RP-1 engine. Some of these same design philosophies can be seen in SpaceX's Merlin. They also use a gas generator cycle, and the first version used an ablative bell. Also they started out with a lower thrust, ISP, and chamber pressure.
All good points. Everything looks better in hindsight. And TR-107 and RS-84 weren't even in development during the EELV competition I don't think. RD-180 existed and RS-68 I think was a left over from some previously cancelled program. So they couldn't have really been considered for the EELV program. But I'm just saying imagine if LM and either Aerojet, TRW, Northrop-Grumman, or PWR develop a domestic engine for Atlas V (instead of carrying over Atlas III's RD-180), they may have very well come up with something like a TR-107 or RS-84 a bit earlier, as LM already had access to RD-180's. So we understood the RP-1 staged combustion concept the Russians had been using for a couple of decades prior. RS-84 came out of knowledge PWR gained from RD-180, as I understand. Not sure if the TR-107 did too.
I think the successful Atlas III showed the capabilities of a high performance kerolox booster with a high performance hydrolox upper stage. So despite the higher ISP of even an inefficient hydrolox engine like the RS-68, the example was there for all to see that RP-1 can make a pretty good booster depsite it's ISP.
Falcon 9 showed us that keeping it simple, even lower isp GG kerolox can be very effective too. The F9 v1.1 has about the same GTO performance as Atlas V-401, despite it not having a staged combustion booster, or a hydrolox upper stage. And the two LV's are roughly similar in size. F9 isn't hugely larger or anything.
Had Columbia not happened, SLI probably would have continued. IF RS-84 or TR-107 had come out of that, perhaps ULA and USAF would have looked to switch Atlas too it as a domestic engine. I dare say SLI and OSP if allowed to continue would have come up with a better STS successor than the sudden switch to CxP did.
If Atlas were using TR-107 or RS-84, there wouldn't be an issue with RD-180's, and the emergency of SpaceX would simply mean ULA would be dropping Delta standardizing on Atlas only, and not have to develop a new LV.