On the verge of ending this update, it is important to make a lengthy digression about the so called "spirit of the Trompowsky".
Indeed, asking for the 2...e6 Tromp line we are examining, an enthusiastic subscriber wrote to me a couple of weeks ago in these terms: "Like GM Peter Wells in his book I firmly believe that if 3. e4 is unplayable against the Tromp, the Tromp as a whole will be a lot less "scary" for Black."
Well, this is not my experience and for me playing the Tromp is a result of a slow and long process. I started chess with 1.e4, then moved on to the 1.Nf3-1.d4 complex in 'Kramnikian' style, then back to 1.e4 with some sharp, up-to-date, heavy opening preparation after which I became an IM without any training guidance. Taking up 1...d5 against 1.d4 or 1.Nf3 I acquired the necessary maturity to become a GM some years later. Then I went back to 1.Nf3 (which is less compromising) when, for certain other professional reasons I had a lot less time to devote to chess.
Next, I went into the magic World of 'd-Pawn specials' o) afterwards with 1.d4 2.Nf3 3.Bg5 and afterwards 1.d4 2.c3 3.Bg5, which is less energy consuming than anything else in terms of getting something solid 'to get your teeth into' out of the opening against good opposition.
As it was with reversed colours, the kind of stuff I had played most of my career, which is also playable without any specific opening preparation! Finally the Tromp which in my opinion gives less room for Black to adapt then interposing any 2nd white move. Thus my approach to the system is different than those players who have adopted it for its surprise value at the beginning, more compact theory nowadays, or its 'scary' aspects. As far as I am concerned, the Tromp is just something entirely natural! like 1.Nf3 d5 2. g3 (or 2.d3, 2.b3) 2...Bg4 with which I have had excellent results for 15 years. Classical chess recommends moving the knights before bishops but it is only because the knights find their development place (Nf3 in case of 1. e4 in 90% of the cases) more easily than the bishops!
There are many examples where the evolution of theory has proved the relativity of this principle. Particularly in the case of 1.d4... I find 2.Bg5 even quite logical, when, against a knight on f6, g5 will most certainly be the square exerting the biggest pressure on the opposing position, and is immediately justified in case of 2...e6 ...! 
Under these conditions, I am sorry but I just cannot consider a move like 3.e4, sacrificing material (the pair of bishops) and forcing White to play dynamically at the very beginning of the game, to be good. At least is it not the way I like to play chess, putting all my eggs into one basket on move 3. In addition, this goes against the predominance of static factors in the modern game that the evolution of theory relays. So I compassionately understand the people who have been put off from playing the Tromp after 2...e6 3.e4?! very well. 
In conclusion, There is no Tromp spirit. It is a delusion in the world of a unique opening! With its subtleties and its completely different ways of handling the position according to Black's second move, but where the general rules of Chess do still apply. The deeper you get to know it, the more you will understand the considerable links it entertains with the 1... d5 (1...c6 2...d5) sphere more than anything else...