Not an user of R or quantmod but typically the batch getters only get the last day, a big problem if you want dividend and split adjusted data. Also, it depends on the connection. My data query (not Yahoo) for about 5000 symbols per day can take 2 hours from Asia (a decent connection) or 5 minutes from a nearby VPS in New York.
This morning my R script ran at its usual speed. I hope the problems yesterday morning were one-off. Thanks to everyone for their suggestions. To answer the question above: for #symbols = 493 Time elapsed(s) (total,per_symbol): 825.32 1.674077 Daily data is retrieved for each symbol from 1990 on, although the average starting year for the symbols may be about 2000.