Welcome to the wonderful world of analog tape!
In all of these conversations, it would be most helpful to have some information about the test tapes being used. There are a least two distinct flavors of CCIR EQ that have been used over the years, so I would expect to see some variations depending on the calibration tapes. In this realm, I have generally found the MRL and AGFA/BASF tapes to be the most consistent, even if they don't match exactly in either azimuth or frequency response. I am making the assumption that a good quality external voltmeter is being used for the reproduce alignment as well. Do not trust the meters on the machine to have flat response! (Although some are fairly decent)
I have compared various test tapes from MRL, BASF, STL, Norton and Ampex from over the years, and have seen significant variations between all of them. Also, as Paul points out, many decks (even those with constant tension), will exhibit significant variations in the actual dynamic "skew" of the tape as it passes across the heads. I remember tearing my hair out over this issue early in my career when working with Ampex 351 and 440 decks (as well as the Scully 280's). You could align the reproduce perfectly on a machine with one tape formulation, but the minute you changed to a different stock, the dynamic skew (and hence the HF response) would change. It doesn't matter if the tape is the same thickness-it has to do with factors related to the oxide and back-coating friction, the actual tape width (with Ampex tapes being slit slightly narrower than 3M, AGFA/BASF, Maxell and others), as well as the base formulation. On many decks (even those with constant tension), you can actually measure the difference in where the tape rides through the head stack and pinch roller. In decks without constant tension control, you could see very significant differences. All of these factors will have a profound effect on the HF response and inter-channel phase relationship of the reproducer.
It is because of these factors that at one point our studio spent a huge sum of money to have Jay McKnight of MRL make a run of tapes on something like 4-5 different stocks, so that we could use the same oxide for the repro alignment that was being used as raw stock for recording.
Although most Studer machines and the Ampex ATR's (as well as the 3M series 79 and Technics RS-1500 series) generally have fewer problems in this regard, I have still encountered issues relating to the difference in dynamic skewing on all of these machines (I use a modified RS-1520 in my home system, mainly because I don't have room in the rack for a larger deck. I still have to tweak the PB azimuth for different oxide formulations, though).
A major factor to keep in mind here is that different tape formulations will establish different head wear patterns as well, so once you run in a headblock with one oxide, and then change it, you will frequently encounter issues with spacing loss across the heads due to the fact that the tape is not contacting the head curvature in the same manner. To make matters worse, if the head tilt or rotation was wrong when the heads were first aligned, you now have an asymmetrical wear pattern to contend with. (We used to encounter this a lot when switching between Ampex formulations and others, due to the slightly stiffer base formulation used by Ampex in the seventies and eighties). If you have a different base material (acetate or polyester), the issue is even more profound.
In addition, I have seen variations in the handling of the tape on machines where the pinch roller formulation has been changed, which is often overlooked as a factor. (Even the Otari MTR-90 machines will exhibit differences in tracking with different rollers).
So, what is novice R-T-R fan supposed to do in the face of all this? The best suggestion I can make is to find somebody in your area that has experience in R-T-R machine alignment. The second would me to make an investment in a MRL IEC curve calibration tape (catalog #21J203 is the one you need). If you are spending $2500+ for a music library, spending $70.00 for a calibration tape well worth it.
For those who are somewhat technically inclined, a good quality (preferably analog) VTVM is also a good investment. You can pick up a working Hewlett-Packard voltmeter on eBay for under $100. The calibration on it will run you maybe $100 to $175. Still a good investment. A basic 'scope is also a must for head alignment. Again, a decent Tektronix or B&K 'scope can be had for anywhere from $150 to $400, depending on condition (I just picked up a good quality 4 channel Tek scope for $175). If you want something that looks cool as well, and have the $$ to spend, you could opt a McIntosh MI-2 or MI-3 Maximum Performance Indicator instead. It will allow you to make the proper phase alignment, and is fun to watch in the dark!
However, if all this is just Greek to you, start checking around your area for a good tech. Expect to pay him (or her) anywhere from $75.00 to $120.00 per hour for their time, depending on their level of expertise, and the equipment they have available.
Perhaps a section could be devoted to what resources are available in various areas of the country in this regard. Also, remember that not all techs will have the experience (and tooling) to work on all machines.
Apologies for the long post...