In Skelcher’s article he goes into some detail on the technique to determine uranium content using a Beta sensitive Geiger counter, including calibration against known batch compositions. He mentions other more costly or inconvieniant methods, such as chemical analysis or gamma spectrometry, but using the Geiger counter looks to be his preferred method. He doesn’t specify which method was used to analyse uranium content in the glass shown in plate 3, but it seems inconceivable that he showed those glasses without confirming their uranium content with at least the Geiger counter method.
If he was wrong about the uranium content in the glass shown, and his accompanying text explaining problems relying on a uv source, then it would seem he would have to be wrong in his Geiger counter technique that looks to be a foundation of his work. I find this unlikely. He says .“...in some [uranium] glasses, especially those with a high lead content, the fluorescence is so weak that there is an element of uncertainty”, that is a clear statement, one would assume proven by his research.
What I take from the article is that if an item is relatively early, say pre 1940 and has a strong uv reaction (glow) then it contains uranium. If it has a weaker glow then it might contain just as much uranium but it would have to be confirmed or denied using an alternative technique. I don’t doubt that glow from manganese is the cause of weak glow in many glass items but that doesn’t necessarily account for every item with a weak glow.