The aluminum heads have pressed valve seats that sometimes present two problems to machining.
The first problem is the small size and position of the seats. If the seats are not installed in nearly perfect concentricity with the valve port, machining them larger can either break through the seat insert or render it so thin that it becomes a weak point. The head can drop the seat under heat and stress conditions, and that gets very ugly very fast. The seats are made barely large enough to accomodate 2.00/1.55" valves, and while incresing the exhaust valve face size another 0.050" is usually considered minor machining, it can be enough to break through on the small factory inserts.
The second problem is seat retention. They are machined for interference and pressed in place. While in operation, the seats normally only see linear forces from the valve landing on it. While in machining, the seats will be subjected to rotational forces. This may be enough to break the seats free and allow them to spin. Once that occurs, the risk of dropping a valve seat insert is very high. If the head is used (not brand new), the thermal cycling it has had during its service life will make the seat even more subject to spinning out, since the expansion rates of the aluminum substrate and steel inserts are different.
Many machinists who are aware of this will recommend cutting or pulling the factory seats and installing replacements, then machining for larger valves. This can present its own problems, since core shift in the casting process can make a larger seat almost impossible without breaking through the castings into the cooling jackets.
Therefore, while it can be done successfully in many cases, there are risks involved, and in
some cases a "simple" valve enlargment can render a casting completely useless. If you ask your machinist about it, don't be surprised if he/she will accept the job but only at YOUR risk, and will not accept responsibility for any damage. If your machinist is not aware of the risks, be wary.
Iron heads merely have induction hardened areas around both the valve seats, and therefore can be machined for 2.02/1.60" valves or even 2.05"/1.55" valves (so long as the exhaust valve is kept smaller to accomodate both of them in the chamber). However, given the flow characteristics of these heads, increasing the intake valve without the exhaust valve is almost useless. The restriction is not really on the intake sides.
As for the aluminum castings being the "better" heads for modification, that's all relative. I know that many of them have had good work done, and have seen improvements as a result. However, I suspect it is easier to start with an iron head and end up with larger port runners and smoother transitions than with an aluminum head. The valve seat inserts alone make blending the short side radii of the valve bowls a very tricky, risky endeavor, and the result would still be a necessarily larger "hump" of material around the short side to keep the seat insert embedded in aluminum. The iron just doesn't care. I also suspect that removing enough material to yeild 194/80cc finished runner volumes would be dangerously close to hitting water on an aluminum casting. Very easy in iron:
