A few things first.
First, the more I hear of Phils Hokkien, the more it strikes me how incredibly similar it is to TWese on two fronts: vocab. and grammar.
Second, the tone differences might be revealing. On T1, T2, T3, T4, and arguably T6 too, the standing (citation) tone value in "Mainstream Philippines Hoklo" (MPH) is the same as the running ("sandhi'd") tone value in "Mainstream Taiwanese Hoklo" (MTH). Yet there are some contexts where the tone value is the same in both dialects. What I mean is, U'll have a sentence where, in MPH, the citation tone is used on a certain syllable; and in MTH, the running tone is used on the same corresponding syllable. The result is that the same tone contour is used on that syllable in both versions (in the two dialects).
This kind of comparison is easy with MPH vs MTH b/c so many structures, words, and expressions are almost identical btw the two... Whereas Penang Hokkien is way different.
One of the theories I wanna "get at" is that a lot of cross-dialectal bleeding, such as a merger of two tonal systems, has taken place, not just in the Phils, but even in TW. Kids would've heard speakers in two different tonal systems. They would've picked up a mixed tonal system, and then found new ways to interpret the "data" and formulate "tonal rules".
As any non-native Hokkien learner knows, the devil of learning Hokkien is in the details of the tone rules. Esp. (I suspect) in the non-creolized dialects.
Another thing I notice, though, is that MPH is more in line with General Sinitic and Written Chinese.
In this thread, I will number each observation (U'll see) for "ease of reference".
Here we go.
#1 She (the speaker) generally drops her nasalization.
#2 "Ông siansíⁿ, lí teh khòaⁿ--siáⁿ?"
The first three syllables are running, running, citation. In TW it would be citation, light+low, light+low.
The last two syllables are high+falling, light+low -- exactly how a lot of TWese say it! In the MPH system, this is clearly analyzable as citation, light+low. In TWese, though, T3 is only high+falling when running; but if it runs, then the following syllable "generally" must be citation.* To "solve" the data, TWese analyze the final, light+low syllable as being citation T4. Thus, "khòaⁿ siaⁿh". But this may be "artificial". It may really just be citation, light+low, and "borrowed" in its exact contours from a dialect with Coanciu tone (e.g. MPH) ... into a dialect with Ciangciu tone (e.g. MTH).
My guess is that it's a sensitive topic for most young spkrs and many middle-aged spkrs, b/c they've been losing the language -- although they will never admit it -- and the elders are not happy about it. And, ironically, one thing the young people have been doing to "appease" the elders has been to study Mandarin.
Alguna otra idea? 
)