First, I'm glad Amazon as the seller is shipping you a new one. That is NOT normal. Maybe Huge Hugh stepped on the package or tossed it over a few rows to another bin during stocking.
Any manual read will have some trade offs. They are, however, the ultimate in trying to measure rain. The tippers will always be subject to heavy fall, etc. But if you don't read often during a hot day, will you loose some from the manual to evaporation? Yes, and some have gone so far as to try putting a little oil in there to float on top and seal it, sort of like the waterless urinals in the parks.
Why would you use it only a few times a year? Put it out and leave it out for the whole season. I wouldn't trust it to freeze, though.
If you've ever visited the desert southwest or been to Cape Canaveral, you'll see what sunlight does to everything, paint, boards, people, you name it.
I'm with the one commenter who said that the bigger the throat the less likely there is for outsplash. But then, too, strong winds will skew the total in some varying ways, too. You'll sometimes see the circular fence of hanging panels around some airport manual read stations.
But then, who knows what the accurate amount is, either? I've seen it rain very hard on one side of a farm field, and the other side get little, so variations over even a few rods will make any worry about the real rain not be worth worrying about if you do your best.
I'm happy with an antique Taylor copper 4" (or 3.5, I forget) that was from the early 1900s. It has three spike feet for trying to get level. And the ends of the throat has a very keen edge, not quite sharp but close enough. But one year I lent it to a friend to play with as he tried to measure various tippers for accuracy (all within a few yards radius of one another in an ideal flat surface, no trees, etc setup) and with it he had one of these as you are playing with, my Taylor, a couple of other manual types of the names I forget, and one storm (after many that were within a couple hundredths of each other, there was over a 1/10th of an inch variation. Surprised the heck out of him. No hail, measured right after the storm passed.
Still a mystery, but I guess I'm not too worried about variation. And yeah, considering that insulators on old telegraph and telephone poles can withstand almost anything less than a .22, I think that the right glass formulation would be workable. Not the kind of glass that they use in the old fashioned yard raingauges that farming companies used to give away.