It's interesting that they decided to make an entire new class table for each multiclass just to get around the problem of frontloading. I guess it's all a little too vague right now to really make judgments, but it sounds pretty convoluted to me. I prefer to use the normal class levels for multiclassing, especially if all they are really worried about is what happens at level one.
I think 5e will have to walk a very careful tightrope because you don't want taking a level in a new class to give too much or too little. Unfortunately, it appears that simplicity (or alternately, rules elegance) will be the casualty.
I've always been of the view that one of the key reasons why spellcaster multiclassing was problematic in 3e was because gaining one spellcaster level led to gains in three or four dimensions:
1. An increase in the number and power of spell slots available (this can be further decomposed into the total number of slots and the power of each slot);
2. Access to more powerful spells, about every other level or so; and
3. An increase in the power of individual spells.
The 4e solution (at least for the AEDU classes) was to equalize the number, strength and access to powers for all classes (although you had to spend feats to swap powers), and tie the strength of individual powers to the level of the power instead of directly to the level of the character.
If 5e multiclassing is going to be largely based on the 3e approach, I think the following tweaks will need to be made to avoid revisiting the issues that some players encountered:
1. The power of a spellcaster's spell slots should be tied to character level. A 10th-level character, even if he has only one level of wizard, should still have access to 5th or, at the minimum, 4th level spell slots (although not necessarily 5th or 4th level
spells - see below).
2. The number of spell slots a spellcaster has access to can depend on the number of levels he has in spellcasting classes. Together, this means that a 10th-level character with just one level of wizard might have access to just one spell, but it's a spell that is worth using.
3. Based on what we have seen so far, the effect of a spell will be fixed based on its level and not the level of the caster. What we have not seen is whether the effect of a spell can be increased by preparing it in a higher spell slot. This second point is quite important because, if implemented, it would go some way to allowing even basic spells to remain relevant at all levels of play.
4. If basic spells can remain relevant at all levels of play, then we can make access to higher-level spells dependant on spellcaster level instead of character level. So maybe the Fighter 5/Wizard 5 can't cast
cone of cold like the Wizard 10 and has only half of the spellcasting endurance, but his
fireball is nearly as good. The Fighter 9/Wizard 1 can only cast a single
burning hands, but it packs a punch when he does.