"Starch" is a very broad term that includes a wide variety of related chemical structures, not only amylose and amylopectin, but minute amounts of associated lipids, proteins, and phosphates as well. Furthermore, the molecular weights and solubilities of the polysaccharide components of starch can vary quite a bit. In short, "starch" is not very well-defined in terms of molecular structure and Mw, at least not for the purposes of studying enzyme specificity. Maltopentaose represents a clearly defined substrate with a known structure and known Mw, and can yield much more useful information regarding enzyme specificity, kinetic parameters, etc. For modeling purposes, especially, having a low-Mw well-defined substrate is especially important. Larger malto-oligosaccharides can also be used, but maltopentaose is the largest that is readily available at reasonable prices. Smaller ones may not entirely fill up the binding site, so may be less useful. Maltopentaose is a good compromise.
The kinetic parameters (Km, Vmax, Kcat) of an enzyme are determined specifically for a given substrate and cannot be adapted to suite another substrate as the enzyme kinetics is bound to change with respect to substrate. For amylases which have broad spectrum of substrates including starch, amylopectin, dextrin etc, it is scientifically wise to carry out substrate specificity studies on the particular amylase using different substrates and then determine the kinetics of the enzyme with respect to each substrate.