"Courts within a given District and Circuit are bound by precedents within their own Districts and Circuits, respectively."
Also:
http://scholar.google.com/scholar_case?about=101740948891558...
I'm pretty sure I didn't make a mistake here.
The decision of one district judge does not bind another district judge in the same district. In the Mariel Boatlift cases, the U.S. District Court for the Southern District of Florida sat en banc precisely to avoid the problem of inconsistent decisions.
The decisions of one federal appellate circuit are not binding on other federal circuits. There are frequently conflicts, even amongst panels of the same circuit. These are first resolved by the circuit court sitting en banc to resolve the conflict. If there are conflicts among the circuits, these are resolved by the Supreme Court. The Supreme Court is not obligated to take these cases and sometimes the conflict is resolved by Congress passing legislation adopting one circuit court's view over another.
These are ancillary and non-essential points to your main argument, but are easily corrected. Pacer can be made better, (criminal cases are generally not included; why?) but lawyers simply do not need Pacer to research substantive or procedural federal law.
From what I've seen, research is done at all levels, though the higher the level, the more likely the citation. For example, district judges like to cite other district decisions in their same courts for routine boilerplate, such as the conditions when a 12(b)(6) motion is appropriate.
I have conducted legal research with PACER multiple times in the course of handling litigation. One key time was when I had to look up ALA v. Pataki--a district decision--which has key implications in the FaceCash case because it deals with the right of states to regulate the internet. It never got to the appellate level, but it's the law of the land now. So I would disagree with your statement that "district court cases are not precedential," because they sometimes are. Many other times I wanted to see how particular motions were drafted in other cases (having never written them before), which is something that you can really only do if you have access to the PDFs of the motions themselves.
Google Scholar does not contain district case dockets or documents at this time, though it is a great resource. LexisNexis is barely usable for novices, expensive, and simply not a realistic option for many people involved in litigation who aren't lawyers.
Even though appellate courts obviously track cases with numbers, those cases are not directly linked from PACER such that you can follow a case backwards or forwards with one click. I've used these systems pretty extensively. I stand by what I wrote.
Aaron