Once you consider the need for the "2-hour rule" for blood values to return to normal pre-race values, the expected impact of several weeks at altitude, and for weeks after the subsequent return to sea-level, and all the things that can go wrong when you don't use the same standard to analyze and compare 2 independent samples, there is really no strong indication or basis from these measurements that anything was necessarily abnormal in her blood due to blood doping. Your hunt for patterns of the unpublished scores would also be invalid before 2009, before blood collection, transport, storage, and analysis standards were in place.
After all that, you may still have your doubts, and I may be unable to change that, but I think part of your doubts should include how much blood doping helped, or can help, elite marathon performances in the first place. Recall no other blood dopers came close to Paula's times in the height of the EPO-era in the early 2000s. When the ABP was introduced in 2009, Ndereba was still #2, some 3m22s behind. No one beat Ndereba until Oct. 2011. As late as April 2017, Paula still had the top-3 marathon times, and still only 2 women had beaten Ndereba, and then only by 27 seconds, and 10 seconds. Fast forward to the present, in the supershoe era, and Paula's #1 world record is now #7, and her #3 performance is now #33. Were it not disqualified, Shobukhova would be #69, and Nderaba's #6 would be #101.
Regarding these two points:
1) Again, Paula was opposed to the witchhunt nature of being pressured to publish the values, and the dangerous precedent it would set for other athletes, due to the high risk of misinterpretation from a public looking to confirm suspicions fueled by sensational journalism.
2) Did she backtrack on altitude for these other two points? I don't recall that, or that her known movements did not corroborate that. I recall her initial statement contained all the factors: recent stays at altitude for the three samples, blood samples drawn too soon after a race for two of the samples, and other reasons why pre-2009 comparisons are unreliable, such as inconsistency between different analyzers, or even the same ones over time. I also recalled she later clarified, that while she had been at altitude before all three samples -- altitude was a primary factor for the 2012 sample, but not the primary factor for 2003 and 2005 samples, which were taken too soon after an intensive race. I haven't seen (and Paula wouldn't have either) how many weeks it takes blood values to return to pre-altitude levels after the descent back to sea-level. But we also know that Paula slept in altitude tents, which would have delayed the return to normal.