JOHNSON. “Whether, indeed, we take [Goldsmith] as a poet,—as a comick writer,—or as an historian, he stands in the first class."
BOSWELL. "An historian! My dear Sir, you surely will not rank his compilation of the Roman History with the works of other historians of this age?"
JOHNSON. "Why, who are before him?"
BOSWELL. "Hume,—Robertson, or the foppery of Dalrymple. Will you not admit the superiority of Robertson, in whose History we find such penetration—such painting?"
JOHNSON. "Sir, you must consider how that penetration and that painting are employed. It is not history, it is imagination. He who describes what he never saw, draws from fancy. Robertson paints minds as Sir Joshua paints faces in a history-piece: he imagines an heroick countenance. You must look upon Robertson's work as romance, and try it by that standard. History it is not. Besides, Sir, it is the great excellence of a writer to put into his book as much as his book will hold. Goldsmith has done this in his History. Now Robertson might have put twice as much into his book. Robertson is like a man who has packed gold in wool: the wool takes up more room than the gold. No, Sir; I always thought Robertson would be crushed by his own weight,—would be buried under his own ornaments. Goldsmith tells you shortly all you want to know: Robertson detains you a great deal too long. No man will read Robertson's cumbrous detail a second time; but Goldsmith's plain narrative will please again and again. I would say to Robertson what an old tutor of a college said to one of his pupils: 'Read over your compositions, and where ever you meet with a passage which you think is particularly fine, strike it out.' Goldsmith's abridgement is better than that of Lucius Florus or Eutropius; and I will venture to say, that if you compare him with Vertot, in the same places of the Roman History, you will find that he excels Vertot. Sir, he has the art of compiling, and of saying every thing he has to say in a pleasing manner.
--Boswell, Life of Johnson

What Goldsmith comically says of himself is very true,— he always gets the better when he argues alone; meaning, that he is master of a subject in his study, and can write well upon it; but when he comes into company, grows confused, and unable to talk. Take him as a poet, his Traveller is a very fine performance; ay, and so is his Deserted Village, were it not sometimes too much the echo of his Traveller. Whether, indeed, we take him as a poet,—as a comick writer,—or as an historian, he stands in the first class." BOSWELL. "An historian! My dear Sir, you surely will not rank his compilation of the Roman History with the works of other historians of this age?" JOHNSON. "Why, who are before him?" BOSWELL. "Hume,—Robertson, or the foppery of Dalrymple." BOSWELL. "Will you not admit the superiority of Robertson, in whose History we find such penetration—such painting?" JOHNSON. "Sir, you must consider how that penetration and that painting are employed. It is not history, it is imagination. He who describes what he never saw, draws from fancy. Robertson paints minds as Sir Joshua paints faces in a history-piece: he imagines an heroick countenance. You must look upon Robertson's work as romance, and try it by that standard. History it is not. Besides, Sir, it is the great excellence of a writer to put into his book as much as his book will hold. Goldsmith has done this in his History. Now Robertson might have put twice as much into his book. Robertson is like a man who has packed gold in wool: the wool takes up more room than the gold. No, Sir; I always thought Robertson would be crushed by his own weight,—would be buried under his own ornaments. Goldsmith tells you shortly all you want to know: Robertson detains you a great deal too long. No man will read Robertson's cumbrous detail a second time; but Goldsmith's plain narrative will please again and again. I would say to Robertson what an old tutor of a college said to one of his pupils: 'Read over your compositions, and where ever you meet with a passage which you think is particularly fine, strike it out.' Goldsmith's abridgement is better than that of Lucius Florus or Eutropius; and I will venture to say, that if you compare him with Vertot, in the same places of the Roman History, you will find that he excels Vertot. Sir, he has the art of compiling, and of saying every thing he has to say in a pleasing manner. He is now writing a Natural History and will make it as entertaining as a Persian Tale."
Boswell: Life of Johnson