Whys and Wherefores

The following thorough and intriguing write-up was submitted by a Dikdukian reader:

It is difficult to research a rule that apparently has no name.

I refer to what some colloquially refer to as the "למה" rule, because it is perhaps best illustrated with a handful of examples in which למה is used twice in close succession, once מלעיל and once מלרע (e.g., Shemot 5:22; Shemot 32:11-12 -- you know it from ויחל on a תענית צבור). The rule is that an otherwise מלעיל word becomes מלרע before a word that begins with an א, ה, or ע (of course, HaShem's name spelled with a י is pronounced with an א and therefore triggers the rule).

R. Yaakov Kamenetsky refers to the rule often in his ספר אמת ליעקב but generally as "that rule I have cited many times." Paul Jouon, in his well-regarded grammar written with T. Muraoka, calls it "hiatus" (a not uncommon grammatical process) but admits the name is less than ideal.

None of this sheds much light on the details of the rule itself. Jouon does mention that the word in question must end in an open syllable, but are there any other environmental factors?

Consider these examples (or nonexamples): 

1. Bamidbar 11:8: שָׁטוּ֩ הָעָ֨ם וְלָֽקְט֜וּ... There is some disagreement about this one, as indicated in many חומשים that mark it as מלעיל but other sources, including מנחת שי and אמת ליעקב, that say it follows the rule and is מלרע. Those who say it is מלעיל claim that that is true to the meaning of the word, but we see other instances where the rule overrides that consideration (e.g., זָד֖וּ עֲלֵיהֶֽם Shemot 18:11). Nevertheless, it is unusual to see Koren and others mark מלעיל if it should be מלרע. My thought was that perhaps the rule is blocked with a תלישה (big or small), as seen in ...
2. Bamidbar 16:7: וְשִׂימוּ֩ עֲלֵיהֶ֨ן קְטֹ֜רֶת where ושימו is marked (in every חומש I've seen that double-marks the תלישה in such a case) as מלעיל. But is טעם part of the environment that blocks the rule? Consider also ...
3. Shemot 25: 24: וְצִפִּיתָ֥ אֹת֖וֹ זָהָ֣ב טָה֑וֹר... R. Kamenetsky says it is מלרע because of "that rule." Conveniently, the word doesn't show up elsewhere (without an א afterward) for comparison, supporting R. Kamenetsky's view by default. But what of other ל-ה verbs that don't trigger the rule, such as the very common ועשית? I had surmised merely that a ל-ה verb doesn't trigger the rule, and thus that the שורש of וצפית is not צפה as otherwise expected. And, in fact, other ל-ה verbs like וראית are still מלעיל before an א (Shemot 33:23, Devarim 4:19, the latter with a תלישה, fwiw). Further, the problem for the other view is that is doesn't explain why ועשית is often followed by an א and yet remains מלעיל (as in Shemot 27:1, וְעָשִׂ֥יתָ אֶת־הַמִּזְבֵּ֖חַ, and we see וצפית as מלרע followed by את in 26:29). The other question here is possible interaction between this rule and that of וי"ו ההיפוך in עתיד/צווי -- but clearly R. Kamenetsky is saying the rule occurs here without saying why it doesn't in other seemingly like situations.

So what are the other environmental conditions that trigger/block the rule? I might like to write down each instance the rule shows up to see what's missing, but perhaps someone has done that already or has the software to find out quickly...?