1. What is the coefficient of friction of a material A ?
Friction and wear are system properties, which means that they may depend on the whole system of materials and mechanics that are in contact with each other, and also on the working conditions and the environment.  So unfortunately, there isn’t a single value of coëfficiënt of friction for a material.  This does not mean, that a coefficient of friction cannot be measured, in any system where the normal load and the resulting friction force are known, a coefficient between both can of course be be calculated.  But it will only be a meaningful number when some conditions are met : did we measure the friction force under realistic conditions (i.e. correlating with your system or application ?), did we measure the friction force over a wide enough range to establish if it's constant or not, did we measure the friction force for a long enough duration to establish whether it's stable or not...
Besides, material properties vary for real engineering materials, so it is likely that there is variation in the coëfficiënt of friction even when comparing nominally identical material coupons.
2. Isn’t the coëfficiënt of friction a fixed number ?
This is a widespread misconception, based on initial experimental work of Leonardo (Da Vinci, not Di Caprio) who found that the friction force of a big block of material was proportional to its weight (normal force) and that the coefficient didn't change when changing the size of his blocks.  These observations were later confirmed and turned in to 'laws' by Coulomb.
But these laws are only valid within a limited range (sizes of the materials, forces, speeds, etc) and only for dry friction.  Most industrial systems today do not operate under the same conditions as Leonardo's simple experiments, so coefficient of friction may vary widely for the same materials.  Just rub your hands when they're dry (warming up) or wet (sticking) or when lubricated with soap : very different friction, isn't it ?
So it only makes sense to talk about 'a' coefficient of friction if it has been experimentally measured in a meaningful way (correlating to your application), repeatable and consistent.

From the above answers, it may be clear that it is not straightforward and obvious to extract a value for a coefficient of friction or to use tabulated values.  If you are designing a part, and need to use a friction value, you may be interested to know that the coefficient of friction of steel-on-steel , even unlubricated, may vary from 0.1 to more than 1.  So which one to choose ?

3. What is the wear of material A vs material B ?
The 'wear' of a material has to be measured also in correlation with the intended application.  For the same reason why there isn't a single coefficient of friction for a material, there isn't a single wear number of wear rate.  In fact, the few standardized test methods that describe how to measure wear resistance of a material, always use a fixed procedure and even fixed test equipment, to be able to publish a single 'wear number'.

4. Can we extrapolate a test with the reality ?
Yes and No.  The No portion is the disappointing - or challenging- part of tribology testing : any lab test will be some simplification of the reality and as such does not replace reality.  There may always be factors in real life, that can not be simulated in the lab, or that are maybe even unknown at the time of testing.
But there is also a YES portion : with the proper know-how it is possible to design meaningful lab tests that will approach reality.  And although they will not be a 100% safe extrapolation, they do deliver a lot of data that are often difficult to obtain with a real life test.  In lab tests, more data can be monitored and stored than in many field tests, so more knowledge about the materials and systems can be collected.  And this knowledge, coupled with experience or modelling, can lead to fairly confident extrapolation to real life conditions.  The message is : the lab test must be the right one !

5. Can you calculate the result of test method A from a result of test method B ?
Usually not.  Because friction and wear are system properties, changing the test method means changing the system and thus the values can and probably will change.  If there are correlations between test methods, they are usually empirically established, or based on the fact that very similar wear mechanisms are being testing in both methods.  But it is not the rule.  It is even dangerous to use an inappropriate method (e.g. method A) to predict the behaviour of materials that should be tested by another more appropriate method (e.g. method B), because you may draw contradictory conclusions with desastrous consequences for your application.

6. Aren’t there any material friction and wear data bases ?
Very few and not easily accessible, and besides very incomplete.  The reason for this is simple : because there are too many applications, each with their own set of conditions and because there are too many materials, it is economically impossible to test them all and store the information in a database.
It is even an economic challenge to do enough repeats in wear tests (see FAQ about statistics !).  Because material properties may vary, and because friction and wear mechanisms are highly dynamic, a minimum of repeats has to be used to construct databases.

7. Can we perform a single test ?
You may want to just explore some different materials, and a single test result will do ?  Yes, of course you can, if you are doing a standardized test that has already been used many times with your class of materials, and the typical variation in test results is known.  Assume you do an extreme pressure test by the ASTM D3233B norm.  Then you know from the standard, that a typical result variation is 250 lb.  If you find that lubricants differ more than this 250 lb, you may conclude that they're significantly different.  But keep in mind history and statistics !  The precision of this standard has been established many years ago, with the products of that era.  Are they still relevant ?  And the 'variation' in the standard refers to a 95% confidence interval.  One result in 20 may still fall outside this bracket.  Outliers are very common in friction and wear testing, due to the complexity of the testing, the mechanisms and the variation in materials properties.  It is safer to plan for at least duplicate testing in any case, and certainly when exploring a new method and new materials.  Because duplicate testing doubles the test time, the test price rises of course, which is often considered a loss/cost.  But just as companies insure themselves against many external and internal risks and damages, duplicate or triplicate testing should be considered as an insurance against wrong decisions.