default objective function fails with cryptic errors on abnormal datasets

[josephguillaume]

With short and sub-daily datasets, "r.sq.monthly" cannot be calculated and results in cryptic errors, e.g. In nseStat(aggrQ, doaggr(X), ref = aggrRef, ...) : "merged time series have no data; incompatible times?"