1.presumably if all loans were paid back, there would be no money in circulation.
No if all loans were paid back there would still be money in circulation because part of base money is in the forms of coins and notes produced by mint and bureau of engraving and printing under treasury. Those money would not disappear even if all loans would be paid back.
Fed balance sheet every expected to be zero (i.e. no outstanding loans)? and what happens to the money in circulation from defaulted loans?
No, as long as the government does not decide to switch to some other monetary system and given the relative efficiency of fiat money that is unlikely to happen.
what happens to the money in circulation from defaulted loans?
That depends on what sort of default are we talking about. If for example government declares it won't pay portion of its debt to Fed then nothing will happen to the money in circulation - the money supply increase becomes permanent (in a sense it cannot be anymore reduced by paying down the debt).
- the "value" of money is presumably based on supply and demand. this presumably counters the idea that the more money in circulation reduces its value because as the population and economy grow, there is a greater need for money, maintaining its value.
This is true but that implications are not completely correct. One of the most simplest model to describe the money market (e.g. supply & demand for money) is as follows (see Mankiw's Macroeconomics 8th ed pp ):
where $M$ is money supply, $V$ is velocity of money (how much is one dollar used), $P$ is price level (change in which gives you inflation/deflation - i.e. increase/decrease of value of money) and $Y$ is the real output. In more complex models of money market people's expectations play role and the relationship is not necessary proportional but this simplified version is useful as a didactic tool.
Here you can see that ceteris paribus (holding $V$ and $Y$ constant) increase in money supply $M$ does increase $P$ (which in turn decreases value of money). You are completely correct that as our economy grows ($Y$ increases) there will be higher demand for money so as long as you increase $M$ just by the same amount as $Y$ increases it would not lead to any inflation. In addition, also $V$ tends to change over business cycle (although in the long term it tends to be constant) which means there is additional leeway for creating new money during some times (although also less during another).
However, it is always possible for central bank to increase money supply way beyond a level that would maintain price stability. That is ultimately just question of how far is central bank willing to go and what is legal (e.g. helicopter money might not be legal every country and there are usually institutional constraints imposed on central banks - e.g. in Europe the legality of ECB monetizing government debt is questioned in some national courts).
This being said modern independent central banks are empirically excellent at maintaining price stability (most are legally required to do so).
do they [these concepts] explain/justify fiat money?
Most of the above would work in other monetary systems as well with some caveats. The main justification for fiat money is that it wastes less resources than comparable systems such as commodity money or gold standard. When you back money by let's say gold, then you are just digging up gold to put it back under ground - wasting scarce resource that has many industrial/commercial uses in electronics, jewelry etc.
Ultimately any money is based on trust that someone else will accept your money and this trust can be created via government requiring all taxes be paid in its currency (you always know the money will be useful for something so you accept it). That is much better than wasting scarce resource in order to create trust (e.g. under the gold standard the trust comes from the fact that you know money is convertible to gold and you know you can always find buyer for that gold because it is sought after commodity).
Additionally, monetary policy under commodity money or gold standard requires for monetary authority to regulate production of that commodity or change the ratio at which money converts to the commodity which is more cumbersome than just having fiat money. So that is an additional reason as well.
what concepts are missing from a more comprehensive understanding?
Explaining comprehensively all monetary economics/macroeconomics is beyond scope of a SE answer. A good macro textbooks that deal with these issues are above mentioned Mankiw Macroeconomics, Blanchard et al Macroeconomics a European Perspective or if you want high level graduate text Walsch Monetary Theory and Policy is very good text.