When there are multiple infractions, you take them in chronological order. In this case, the first infraction was the 3♥ bid putatively based on UI from the alert of 2♥. Did the UI suggest bidding 3♥? Surely. Was there a logical alternative? I don't know; I'd have to see the hand. If there was, and if bidding 3♥ can be demonstrated to have been suggested over the LA (likely, IMO) then I would adjust the score on the assumption that the LA was called, rather than 3♥. Let's say that LA is pass, leaving the pair in 2♠. Now there's no need to consider the second infraction, because in this scenario it will never have happened. TD might still issue a PP for either infraction, though. Generally speaking, in a club game, I wouldn't issue PPs, except perhaps in the form of a warning. But if I warn players, I'm saying "do it again, and I'll issue a PP", and I'm going to carry through with that.
It would be interesting if the 3♥ bidder claimed that his hesitation was due to trying to figure out what his responsibilities were in the presence of the UI from the alert.
Addendum 1: Law 12C1(d) allows the TD to award an artificial adjusted score, but only when "the possibilities are numerous or not obvious". It is rare that this should be the case. This law should not be used by the TD as an excuse to be lazy.
Addendum 2: in North America, Law 12C1(e) applies, and the NOS get "the most favorable result that was likely had the irregularity not occurred", while the OS get "the most unfavorable result that was at all possible". In most of the rest of the world, Law 12C1© applies, and the players get a weighted score on the basis of the possible results. This does not change the principle stated in addendum 1.

Help

(1♣) _P (1♠) 1NT
(2♣) 2♥ (_P) 2♠
(_P) 3♥ (_P) ??
2♥ was by agreement, a transfer. What would you bid?
IMO 4♣=10, 4♦=9 4♠=8, 3N=2
I agree with most of what Cardsharp wrote. I agree with Blackshoe's ruling and his opinion that patner's 3♥ bid is suspect. Admittedly, it is hard to construct a hand consistent with partner holding long ♠ but that was just as true over his original transfer as over his slow 3♥ rebid, Perhaps, the deal on the left is plausible. If that construction is deemed unlikely, then perhaps somebody made a mistake (but here, the unauthorised information makes it likely that the error was partner's). The bottom line is that 4♠ was a logical alternative to the 3N suggested by partner's hesitation.