'Es ware deshalb von grosstem Interesse, bis zu wie groszer Sonnennahe helle Fixsterne bei Anwendung der starksten Vergrosserungen _bei Tage_ (ohne Sonnenfinsternis) gezehen werden konnen.'
My translation:
'It would therefore be of the greatest interest [to know], until how near to the sun bright fixed stars could be observed, under the largest magnification and by day (without solar eclipse).'
One can see venus with the naked eye during the day - but not when it is near the sun.
I expect that a red filter would get rid of some of the scattered atmospheric light. A telescope should be able to see some of the brighter stars during the day.
But to get the scope so near the edge of the sun that you can measure that arc-second displacement, you'd have quite a lot of sunlight entering the telescope. The atmospherically scattered light in that region is quite intense.
One can use an "Occulting Disk" but they are of limited effectiveness. When the sun is in the sky, hold your hand out at arm's length, then block the sun with your thumb; you'll see that the skylight is quite bright around your thumb. That's not from the solar corona - not much of it - but from atmospheric scattering.
Goldberger was quite specific that Einstein asked that the telescope be pointed right at the edge of the sun. I've studied german but not in enough depth that I could translate Einstein's letter.
One arc second is not far from the limit of the resolution of an amateur telescope. While in principle the larger telescopes could resolve finer detail, atmospheric turbulence leads them not to resolve much better - they can see dimmer objects, but not more clearly. It is only in recent years that adaptive optics, faster imaging sensors and more-sophisticated image processing algorithms enable better resolution.
I don't know but expect that when the 1919 expedition confirmed GR, it was done by measuring the positions of multiple stars to determine that their positions were, on the average, out of place. The scope they used was not that big, as they had to carry it out into the African countryside.
The tensor calculus used in GR is quite complex. To reduce that to specific predictions of the star locations would require quite a lot of arithmetic. Back then they would have used logarithm tables and adding machines.
I am willing to concede the Einstein could balance his checkbook, but numerical analysis is hard enough to get right on a modern computer, it is quite hard when done the old-fashioned way, with logarithms and the like. It would be easy to write down a single digit incorrectly, or to not carry the computation out to enough significant figures.
Dr. Goldberg was also quite specific about Einstein's computations - not his theory - being "wildly inaccurate".
Finally, that article brought back many memories for me, some good some bad. I love the Huntington, I love Mt. Wilson but I had a rough time at Caltech.
I was, for a very brief time, a solar astronomer, but at Big Bear Lake.