What were the differences on the LM ascent stage engines on Apollo 9, 10, and 11?

Question

According to Apollo By The Numbers, the LM ascent stage of Apollo 11 ("Eagle", LM-5) had maximum rated thrust of 3218 lbf (14.31 kN), and subsequent missions had very similar figures. The engine itself is usually advertised as producing 3500 lbf (15.56 kN).

Apollo 9's "Spider" (LM-3), on the other hand, is shown as having maximum thrust of 2524 lbf (11.23 kN), and Apollo 10's "Snoopy" (LM-4) as 1650 lbf (7.34 kN).

Why were the LM ascent thrust ratings so dramatically different between Apollo 9, Apollo 10, and the subsequent missions?

I know that the original Bell-designed injector for the ascent engine was causing problems, and Rocketdyne stepped in with a new injector very late in the program, but according to this timeline, the new injector was flown on all manned LM flights starting with Apollo 9's LM-3, and it seems very unlikely that an injector with stability problems would have been flown on a critical manned test flight.

I can think of a few possible explanations:

1. Simple errors in the figures from either Apollo press kits or the transcriptions of the press kits to ABTN.
2. Testing of the Rocketdyne injectors at full power had not been completed, so the rated power levels were fixed at artificially low values. This seems unlikely to me, particularly with LM-3 being more powerful than LM-4.
3. The ascent engine power on LM-4 was deliberately restricted so that the acceleration of the ascent stage with its light fuel load would be similar to that of LM-5. However, the math doesn't seem to work: the loaded weight of the LM-4 ascent stage was more than half that of LM-5, so acceleration would be substantially lower across the entire ascent stage burn.
4. The ascent engine power on LM-4, along with the fuel load, was tuned such that the staging and ascent from the LM's 9-mile periapsis would leave the ship in a similar orbit to LM-5's ascent from the lunar surface. This leaves the LM-3 figure mysterious.

Answer 1

The Apollo 10 press kit itself says the APS is rated at 3500 lbf (1589kg) thrust, and the Mission Report says the preflight predictions and in-flight inferences were in the 3434-3499 lbf range. Actual thrust isn't measured in flight, but inferred from tank pressure regulators in the system. The number 1650 which appears in the Apollo By The Numbers table happens to be the loaded ascent stage oxidizer weight for that mission.

Likewise, the Apollo 9 press kit gives the 3500 lbf rating, and predicted and inferred thrust values of 3338-3508 lbf. Again, the loaded ascent stage oxidizer value, 2524 lbs, is erroneously entered as the ascent thrust value in ABTN.

Therefore, my hypothesis 1 was correct: the rated values in ABTN are simply transcription errors from the respective mission reports for Apollo 9 and 10, and the ascent engine was rated 3500 lbf on all flights.

---

I feel a little bad about taking someone else's bounty for answering my own question -- but not all that bad. ;)